Cytoskeletal remodeling in differentiated vascular smooth muscle is actin isoform dependent and stimulus dependent

Differential ultracentrifugation of homogenates for ratios of filamentous- and globular-actin.

The protocol was according to manufacturer's instructions (Cytoskeleton, Denver, CO) with some modifications. Each muscle strip (∼20 mg wet weight) was placed into a glass homogenizer tube (Kimble Kontes, Vineland, NJ) and quickly homogenized by a motorized tissue grinder in 200 μl of lysis buffer (37°C); 50 mM PIPES (pH 6.9), 50 mM NaCl, 5 mM MgCl₂, 5 mM EGTA, 5% (vol/vol) glycerol, 0.1% Nonidet P-40 (NP-40), 0.1% Triton X-100, 0.1% Tween 20, 0.1% 2-mercapto-ethanol, 0.001% antifoam C with 100 mM ATP and a cocktail of protease inhibitors; and 0.4 mM tosyl arginine methyl ester, 1.5 mM leupeptin, 1 mM pepstatin A, and 1 M benzamidine (Cytoskeleton). Tissue homogenates were placed on ice until all samples are completed and then incubated at 37°C for 10 min. Tissue homogenates were transferred to a prewarmed (37°C) ultracentrifuge and spun at 150,000 g for 1 h at 37°C to separate the globular (G)-actin (supernatant) and filamentous (F)-actin fractions (Beckman, Fullerton, CA). The pellets were resuspended with the same amount of volume of ice-cooled distilled water containing 10 μM cytochalasin D. The pellets were dissolved by trituration with a pipette and left on ice for 1 h, vortexing every 15 min for 1 min to dissociate F-actin. Resuspended solutions were centrifuged at 2,300 g for 5 min at 4°C. The second supernatants (F-actin) were collected, and all samples were diluted with appropriate loading buffer and boiled for 5 min. The samples were stored at −80°C until further separation by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblotting.

Immunoblotting.

Quick-frozen tissue samples were homogenized and analyzed via immunoblotting as previously published (37). Briefly, the homogenized samples were boiled in Laemmli sample buffer. The samples were separated using SDS-PAGE and transferred to a polyvinylidene difluoride membrane (Millipore). Protein transfer was confirmed by staining the membrane with Ponceau S red. The membrane was stained with an appropriate primary antibody overnight at 4°C. After extensive wash with Tris-buffered saline-Tween 20, the membrane was stained with an appropriate secondary antibody. Densitometry of the actin bands was performed with the software for an Odyssey infrared imaging system (LI-COR Biosciences).

Single cell isolation from aorta.

Cells were freshly enzymatically dissociated from aorta tissue by a previously published method (37). Briefly, the aorta was carefully dissected to remove all connective tissue and the endothelium under a dissecting microscope. About 100 mg (wet weight) of aorta tissue was then cut into 2 × 2-mm pieces and placed into a silicon-coated flask containing 7.5 ml Ca²⁺, Mg²⁺-free Hanks' balanced salt solution (HBSS), ∼100 U/ml type II collagenase (Worthington), ∼1.0 U/ml elastase (Roche Diagnostics), and 0.2% BSA (Sigma). The flask was secured in a shaking water bath (rate, ∼100 cycles/min; and temperature, ∼34°C) under an atmosphere of 100% oxygen for 80 min. The tissue pieces were filtered on a nylon mesh (pore size, ∼0.5 mm) and washed with ∼12 ml Ca²⁺, Mg²⁺-free HBSS, and 0.2% BSA. The wash buffer containing the dissociated cells was collected and poured over glass coverslips on ice under an atmosphere of 100% oxygen. Cells were allowed to adhere to the coverslips for ∼1 h. After the first digestion, three additional digestions were performed with 50% less elastase for ∼20–30 min. After each of these additional digestions, the tissue pieces were filtered and washed, and the dissociated cells were allowed to adhere to coverslips for ∼1 h. The coverslips were washed with Ca²⁺, Mg²⁺-free HBSS, and 0.2% BSA, and Ca²⁺ and Mg²⁺ were then reintroduced to the dVSMCs by gradually adding CaCl₂ and MgCl₂ into the medium surrounding the coverslips until they reached concentrations of 1 mM and 1.2 mM, respectively. In one coverslip of each batch, it was confirmed that the cells were capable of shortening in sucrose-free PSS. Other coverslips were put in a PSS containing 300 mM sucrose, which increases the tonicity of the solution and prevents shortening directly at the cross bridge level (20), allowing use of the coverslips for imaging. Sucrose HBSS has been shown to cause no change in resting or activated Ca²⁺ levels, PKC signaling, ERK activation time courses, cytoskeletal structure, or resting cell length in differentiated smooth muscle cells (dSMCs) (10, 28, 31, 50, 51).

Cell staining and imaging.

All stimulated and unstimulated dSMCs were fixed with 4% paraformaldehyde for 10 min in 100 mM sodium phosphate buffer. The fixative was quenched by washing the cells with 0.1 mM glycine in 1% BSA HBSS. The cell membranes were then permeabilized using 0.1% Triton X-100 in 1% BSA HBSS. Nonspecific binding sites were blocked by incubating the cells with 10% goat serum in 0.05% Triton X-100 and 1% BSA HBSS. After blocking, the cells were incubated with the desired primary antibody diluted in 2% goat serum, 0.05% Triton X-100, and 1% BSA HBSS. Cells were incubated with the appropriate secondary antibody diluted in HBSS. Cells were washed six times with HBSS and mounted with FluorSave (Calbiochem). Cells were costained with DNase I (Alexa Fluor 488; 300 nM) and phalloidin (Alexa Fluor 568; 1:50 dilution) for 20 min. Latrunculin B (30 nM) was used for pretreatement for 45 min. Images of immunostained dSMCs were taken with a Kr/Ar laser (BioRad Radiance 2000) scanning confocal microscope equipped with a Nikon X-60 [numerical aperature (NA), 1.4] oil immersion objective. Images were recorded with Laser Sharp 2000 (Bio-Rad) software. In all cases, it was confirmed that there was no detectable background fluorescence by staining cells in the absence of a primary antibody. In addition, for all colabeling experiments, it was confirmed that there was no detectable cross talk between fluorescent labels by exchanging excitation/emission filters on single-labeled cells.

Digital confocal microscopy for phalloidin-to-DNase I ratio.

Ratio measurements for phalloidin-to-DNase I ratios followed a modification of previously published methods (17). Phalloidin has been shown to be highly specific for F-actin, and DNase I was used to identify G-actin and discriminate it from F-actin (34). Images were acquired using a Kr/Ar laser (Bio-Rad Radiance 2000) scanning confocal microscope equipped with a Nikon X-60 (NA, 1.4) oil immersion objective. All images were digitally scanned at 166 lines/s, Kalman averaged three times with minimal pinhole size, and quantified with Bio-Rad Laser Sharp 2000 for Windows NT. For each cell, the single central nucleus was first located at the widest and thickest part of the cell. In previously published studies, nuclear staining confirmed that visual inspection gives accurate identification of the nuclear location (29). The position of the nucleus was especially obvious in the current study, since the phalloidin staining was characteristically decreased in the nuclear region. Others have also note that nuclear stains give an identical image to that seen in the bright field image of fixed/permeabilized cells (26). The stage was moved so that the nucleus was positioned to one side of the field of view. We used center optical sections and drew each box to extend from cell edge to cell edge. We used three regions of interest (ROIs) per cell in relatively straight tail regions and excluded the nucleus area. In most cases, we covered the majority of the straight, nonnuclear space (Fig. 2A, inset). The average integrated intensity per pixel within each ROI was determined for each wavelength. The ratio of the value for phalloidin was divided by that for DNase I. The ratio values for the three ROIs for each cell were averaged, and the final number was taken as an index of the ratio of F-actin-to-G-actin (F/G ratio) for that cell as previously described (34). One advantage of the phalloidin/DNase I method used here is that the relative abundance of structures too small to resolve by confocal microscopy (such as G-actin monomers) can be quantitated by integrating the fluorescent signal.

Open in a separate window

Fig. 2.

Single cell F/G actin turnover demonstrated by phalloidin and DNase I costaining. A: illustrative cell images stained with Alexa 488-DNase I (G-actin; green; left), Alexa 568-phalloidin (F-actin; red; middle), and merged images (right) in unstimulated, Lat (0.03 μM), KCl (51 mM), DPBA (10 μM), PE (10 μM), or jasplakinolide (Jas; 0.5 μM)-treated dVSM cell (dVSMC). Inset: diagram of cell with white boxes to illustrate how regions of interest (ROIs) were defined. B: quantitative analysis of ROIs stained with fluorescent phalloidin and DNase I in cells. **P < 0.01, compared with unstimulated cells. Average is ± SE for 8–41 cells.

Two-dimensional gel electrophoresis.

Two-dimensional gel electrophoresis was performed following previously published methods (15, 46, 47). Briefly, samples were homogenized in a buffer containing 9.5 M Urea, 2.0% NP-40, 5% β-mercaptoethanol, 1.6% ampholines (pH 4–6.5), and 0.4% ampholines (pH 3–10). The isoelectric focusing gels were cast according to O′Farrell's method (47) except that the ampholyte mixture was restricted to the 4–6.5 intervals [9.2 M Urea, 4% acrylamide, 2% NP-40, 1.6% ampholines (pH 4–6.5), 0.4% ampholines (pH 3–10), 0.01% ammonium persulfate, and 0.1% N,N,N′,N′-tetramethylethylenediamine]. The gels were allowed to polymerize overnight at room temperature and were prefocused at room temperature using 20 mM NaOH (anode solution) and 20 mM phosphoric acid (cathode solution) for 15 min at 200 V, 15 min at 300 V, and 15 min at 400 V. Isoelectric focusing was performed for 3 h at 400 V and 15 to 16 h at 500 V using Hoefer Scientific tube gel equipment. The second-dimension gels were run as a regular SDS-PAGE. The first-dimension tube gel was transferred to the slab gel by running buffer containing 250 mM Tris, 1.92 M glycine, and 1% SDS.

Synthesis of NH₂-terminal peptide of actin isoforms.

Five tetra- or penta-peptides representing the NH₂-terminal actin isoform sequence were synthesized by solid-state peptide synthesis using fluorenylmethoxycarbonyl chemistry. The isoforms, acetylated at their NH₂ termini, were α-skeletal muscle actin, Ac-DEDE; γ-smooth muscle actin, Ac-EEETT; γ-cytoplasmic actin, Ac-EEEIA; β-cytoplasmic actin, Ac-DDDIA; and α-smooth muscle actin, Ac-EEED. All isoforms were synthetically coupled to a TAT protein transduction domain comprising residues YGRKKRRQRRR.

To add a fluorescent label to the peptide, a bridge consisting of the sequence GGG-β-alanine-FITC was manually added as a side chain at the first lysine residue of the TAT sequence. The ɛ-amino group of this residue was protected during the initial synthesis by the acid labile 4-methyltrityl (MTT) group. The MTT was selectively removed from the peptide by five consecutive 1-min treatments with 2% trifluoracetic acid (TFA) plus 3% tri-isopropylsilane, leaving the peptide still attached to the resin with all other reactive side chains still protected. The three glycine residues and the β-alanine were added to this site with capping by acetic anhydride carried out after coupling each bridge residue. The FITC was coupled to the amino-group of the β-alanine. The peptides were purified to >95% using reversed-phase HPLC.

Loading of NH₂-terminal peptide of actin isoforms.

An NH₂-terminal FITC tag was added to monitor loading efficiency, and a TAT sequence (GYGRKKRRQRRR) was added to facilitate loading (1). α-Skeletal muscle actin peptides were used as a negative control peptide. Aorta tissue was loaded with the peptides as following protocol. At the beginning of all experiments, the tissue was challenged with 51 mM KCl-PSS to test for viability and to normalize for differences in the size or strength of each strip. As Kitazawa and his colleagues mentioned in their publications (11, 33, 63), most blood vessels show an increase in the amount of contraction observed to KCl challenges with time during the experiment. The data in this study were normalized to the first KCl challenge at the start of the day, and thus the response to a later KCl challenge was generally greater than 100% of the initial challenge. The KCl-PSS was washed out several times with PSS. The aorta tissues were allowed to relax for 1 h. The aorta tissues were then incubated for 3 h with 50 μM of each peptide in PSS aerated in oxygen. The peptide solution was changed with fresh peptide solution every hour. A parallel experiment without peptide in the solution was used as a sham. After the incubation period, the tissues were stimulated with 12-deoxyphorbol-13-isobutyrate-20-acetate (DPBA; 10 μM, for 20 min), phenylephrine (PE; 3 μM, for 10 min), or KCl-PSS (51 mM, for 10 min). The tissues, after peptide loading and extensive washing, are visually inspected, before proceeding with the experiment, for the yellowish-green color, confirming the successful loading of the FITC-labeled peptide. Details of this method are published in past studies from our laboratory using the same tissue and successfully loading TAT-peptides and peptides with similar TAT-like sequences (17, 24, 37).

Reagents.

Jasplakinolide, phalloidin, and DNase I were purchased from Molecular Probes (Eugene, OR). Latrunculin B was purchased from A. G. Scientific (San Diego, CA). A G-actin/F-actin assay kit (BK 037) was purchased from Cytoskeleton. DPBA was purchased from LC Laboratories (Woburn, MA). Pharmalytes for isoelectric focusing were purchased from Amersham Pharmacia Biotech AB (Uppsala, Sweden). Paraformaldehyde was purchased from Electron Microscopy Sciences (Hatfield, PA). General laboratory reagents used from analytical grade or better and were purchased from Sigma (St. Louis, MO) and Bio-Rad (Hercules, CA).

Antibodies.

The following antibodies were used: rabbit polyclonal G-actin antibody (1:500, 0.5 mg/ml; Cytoskeleton); monoclonal anti-α-smooth muscle actin (clone 1A4; 1:10,000, 4.4 mg/ml) and monoclonal anti-β-actin (clone AC-15; 1:150, 1.6 mg/ml) (Sigma); sheep anti-actin, γ-polyclonal antibody (1:1,000; Chemicon); Alexa Fluor 488 goat-anti mouse IgG (2 mg/ml), Alexa Fluor 568 goat-anti-mouse IgG (2 mg/ml), Alexa Fluor 488 donkey anti-sheep IgG (1:2,500, 2 mg/ml), Alexa Fluor 680 goat-anti mouse IgG (2 mg/ml), and Alexa Fluor 680 donkey anti-sheep IgG (1:2,500, 2 mg/ml; Molecular Probes); and IR-Dye 800 goat anti-rabbit IgG (1.0 mg/ml), IR-Dye 800 goat anti-mouse IgG (1.0 mg/ml), and IR-Dye 700 goat anti-mouse IgG (1.0 mg/ml; Rockland). The concentration of antibodies is in stock.

Alterations of actin polymerization state modulate contractility in a stimulus-dependent manner.

To test the hypothesis that the net increases of actin polymerization described above cause changes in contractility, we pretreated live aortic smooth muscle tissue with either latrunculin or jasplakinolide. As shown in Fig. 3A, steady-state tonic contractile force in response to DPBA is significantly inhibited by pretreatment with latrunculin. In contrast, the average tonic contractile force response to DPBA is slightly increased by jasplakinolide pretreatment, but the difference is not statistically significant.

Open in a separate window

Fig. 3.

Modulation of polymerization regulates contractility. Contractile responses to DPBA (1 μM; A) or KCl (51 mM; B) in Lat (0.1 μM)- or Jas (0.5 μM)-pretreated aorta tissues. Vehicle (V) for Lat or Jas was ethanol or DMSO, respectively. **P < 0.01, compared with vehicle. Average is ± SE (n = 5 to 6).

The above results could simply be the nonspecific result of latrunculin-induced alterations of cellular cytoskeleton content. However, we performed the same experiment in preparations activated by 51 mM KCl-PSS, and neither latrunculin nor jasplakinolide pretreatment has any significant effect on force generation (Fig. 3B). These results also point to a difference in the mechanism by which KCl-PSS increases net actin polymerization, since this must involve a latrunculin-insensitive mechanism, i.e., one that is not affected by the sequestering of G-actin monomers.

It is worth noting that the treatment of aortic tissues with latrunculin or jasplakinolide alone has no detectable effect on intrinsic basal tone, even though the ferret aorta has previously been shown to display an active, intrinsic, basal tone (23). This further supports the selectivity of these effects.

Inhibition of net polymerization by latrunculin is agonist dependent.

To confirm that the stimulus-dependent modulations of contractility caused by latrunculin or jasplakinolide pretreatment are paralleled by changes in actin polymerization, we pretreated freshly dissociated dVSMCs with the same agents as above and quantitated the effect on F/G ratios. Figure 4A shows illustrative images of individual freshly dissociated dVSMCs under different conditions. Although considerable cell-to-cell variability was noticed, when the results were quantitated (Fig. 4B) by ratiometric analysis, it was found that latrunculin alone causes no significant change in the phalloidin-to-DNase I staining ratio. However, latrunculin pretreatment significantly inhibits both PE- and DPBA-induced increases in the phalloidin-to-DNase I ratios.

Open in a separate window

Fig. 4.

Actin remodeling by DPBA or PE, but not by KCl, is inhibited by Lat. A: illustrative images stained with Alexa 488-DNase I (G-actin; green; left), Alexa 568-phalloidin (F-actin; red; middle), and merged images (right) in cells without or with Lat (0.03 μM) pretreatment. B: quantitative analysis of ROI of freshly dissociated cells costained with fluorescent phalloidin and DNase I. **P < 0.01, compared with unstimulated (white bars); ##P < 0.01, compared with Lat-untreated cells (black bars). Average is ± SE for 18–40 cells.

In contrast, when 51 mM KCl is used as a stimulus, pretreatment with latrunculin causes no significant inhibition of the KCl-induced increase in the F/G ratio (Fig. 4B). Thus these results are consistent with the fact that latrunculin pretreatment has no significant effect on the depolarization-induced increases in contractility (Fig. 3) and, furthermore, point to a difference in the mechanism by which KCl depolarization, versus DPBA, causes a net increase in actin polymerization with the former being latrunculin insensitive and the latter being latrunculin sensitive.

Multiple actin isoforms coexist in dVSM.

Whether actin isoforms are differentially distributed within individual smooth muscle cells remains controversial. Thus we first determined the actin isoform subtype composition in intact ferret aorta tissue. Homogenates of intact aorta tissues display three predominant spots at 42 kDa, when run on two-dimensional gels (Fig. 5A). These three spots were confirmed to be α-smooth muscle actin, β-cytoplasmic actin, and γ-actin (from acidic to basic; Fig. 5A) by mass spectrometry (Tufts University Core Facility, Boston, MA). γ-Smooth muscle actin and γ-cytoplasmic actin cannot be separated by two-dimentional gels, but the relative amounts of proteins in the three spots were quantified by densitometry after staining the two-dimensional gels with Proteom IQ Blue. The major actin component in aorta is α-smooth muscle actin (∼66.2% of total actin), followed by β-cytoplasmic actin (∼21.1% of total actin), and γ-actin (∼12.7% of total actin). This result is consistent with a previous study on vascular actin isoform composition from different species (15).

Open in a separate window

Fig. 5.

Actin isoforms in aorta. A: 2-dimensional gel of aorta homogenate stained with Coomassie blue. B: images of illustrative dVSMCs immunostained with isoform-specific actin antibodies. Scale bars, 10 μm. IEF, isoelectric focusing.

To determine whether the subcellular distribution of actin is isoform dependent in dVSMCs, we stained freshly dissociated aorta cells with isoform-specific antibodies for α-smooth muscle actin, β-cytoplasmic actin, and α-/γ-actin (Fig. 5B). The α-smooth muscle actin is present in longitudinally or slightly tangentially oriented filamentous bundles (arrows) in the freshly isolated dVSMCs (top) in a manner similar to that shown previously for myosin-containing contractile filament bundles (50). In contrast, β-cytoplasmic actin was prominent in the cell cortex and in shorter punctuate spots (arrows) throughout the cell (middle). Unfortunately, there is no available entirely specific anti-γ-antibody. We found that the commonly used commercially available polyclonal γ-antibody (Chemicon, Temecula, CA) recognizes, on two-dimensional gels, both α- and γ-actin isoforms (data not shown). When cells are stained with this antibody, the distribution of staining includes some longitudinal filaments (arrows) expected from α-smooth muscle actin cross-reactivity (Fig. 5B, bottom), but a clear discrimination of the γ-actin staining from that due to α-smooth muscle actin could not be determined.

Isoform-specific NH₂-terminal actin peptides depress contractile force.

To further test whether actin isoforms modulate the contractile function of smooth muscle, we synthesized NH₂-terminal decoy peptides of actin isoforms (Table 1) coupled to a cell permeation TAT tag sequence and introduced them to intact dVSM. It has previously been reported that the NH₂-terminal α-smooth muscle actin tetrapeptide inhibits the formation of stress fibers and wrinkles on silicone substrate in an isoform-specific manner in cultured cells (6, 22). The NH₂-terminal actin sequences are also known to include sequences involved in the association with many actin-binding proteins, and since considerable evidence in nonmuscle cells indicates that, despite the large degree of sequence conservation, actin displays isoform-specific functions, we synthesized all of the relevant NH₂-terminal peptides as putative decoy peptides. We also made the α-skeletal muscle peptide as a negative control to test the specificity of the approach.

As expected, the α-skeletal muscle actin peptide has no significant effect on the steady-state contractile force induced by either DPBA or PE (Fig. 6). The introduction of the β-cytoplasmic actin peptide also has no significant effect on either DPBA- or PE-induced smooth muscle contraction. However, the contractile force induced by DPBA is significantly inhibited by the introduction of α-smooth muscle, γ-smooth muscle, and γ-cytoplasmic actin peptides (Fig. 6A). Interestingly, the contractile force induced by PE is significantly inhibited by the introduction of γ-smooth muscle, and γ-cytoplasmic actin peptides, but not in α-smooth muscle actin (Fig. 6B).

Open in a separate window

Fig. 6.

Decoy actin peptides inhibit contractility in an isoform-dependent manner. Steady-state force in the presence of DPBA (10 μM; A), PE (3 μM; B), or KCl (51 mM; C) in the presence of peptides or sham treated (Table 1) is shown. Contractile responses were normalized to force to a challenge with KCl physiological salt solution (PSS; 51 mM KCl) at the beginning of each experiment. *P < 0.05, effects of peptides compared with sham (Sham). Average is ± SE (n = 3–6). A-SM, α-smooth muscle actin; beta, β-cytoplasmic actin; G-SM, γ-smooth muscle actin; G-CYT, γ-cytoplasmic actin; A-SK, α-skeletal muscle actin.

In contrast, in the same experimental protocol, but with preparations activated by KCl-PSS, none of these peptides has any significant effect on force generation (Fig. 6C), further pointing to selectivity in the action of these peptides. There was also no detectable decrease in intrinsic basal tone with any of peptides.