Airway smooth muscle tone modulates mechanically induced cytoskeletal stiffening and remodeling

Linhong Deng, Nigel J. Fairbank, Darren J. Cole, Jeffrey J. Fredberg, Geoffrey N. Maksym

Abstract

The application of mechanical stresses to the airway smooth muscle (ASM) cell causes time-dependent cytoskeletal stiffening and remodeling (Deng L, Fairbank NJ, Fabry B, Smith PG, and Maksym GN. Am J Physiol Cell Physiol 287: C440–C448, 2004). We investigated here the extent to which these behaviors are modulated by the state of cell activation (tone). Localized mechanical stress was applied to the ASM cell in culture via oscillating beads (4.5 μm) that were tightly bound to the actin cytoskeleton (CSK). Tone was reduced from baseline level using a panel of relaxant agonists (10−3 M dibutyryl cAMP, 10−4 M forskolin, or 10−6 M formoterol). To assess functional changes, we measured cell stiffness (G′) using optical magnetic twisting cytometry, and to assess structural changes of the CSK we measured actin accumulation in the neighborhood of the bead. Applied mechanical stress caused a twofold increase in G′ at 120 min. After cessation of applied stress, G′ diminished only 24 ± 6% (mean ± SE) at 1 h, leaving substantial residual effects that were largely irreversible. However, applied stress-induced stiffening could be prevented by ablation of tone. Ablation of tone also inhibited the amount of actin accumulation induced by applied mechanical stress (P < 0.05). Thus the greater the contractile tone, the greater was applied stress-induced CSK stiffening and remodeling. As regards pathobiology of asthma, this suggests a maladaptive positive feedback in which tone potentiates ASM remodeling and stiffening that further increases stress and possibly leads to worsening airway function.

  • stress
  • contraction
  • actin
  • hyperresponsiveness
  • asthma
  • magnetic twisting cytometry

the airway smooth muscle (ASM) cell is subjected to tidal stresses associated with breathing (8). These stresses have been shown in vitro to cause changes in ASM cell function and cytoskeleton (CSK) remodeling. For example, Smith et al. (32–37) have shown that chronic cyclic mechanical stretch (up to 12 days) promotes proliferation, force production, calcium sensitivity, shortening velocity, shortening capacity, CSK organization, and stiffness. Similarly, our laboratory has shown recently that localized mechanical stress can cause acute (within 60 min) ASM cell stiffening and actin accumulation at the sites of stress application (11). These observations, taken together, suggest that mechanical stress modulates CSK remodeling in a way that, in long term, enhances the ASM contractile function.

However, CSK remodeling in the ASM may also be modulated by contractile activation of the cell. It has been shown in fibroblasts, for example, that stress-induced growth of focal adhesions and strengthening of integrin-dependent linkages to the CSK are sensitive to intracellular contractile activation (5, 9). Moreover, it has also been shown in primary cultured ASM cells and fresh tissue strips that contractile activation promotes actin polymerization and remodeling (17, 26). Recently, we found in the cultured ASM cell that both contractile activation and mechanical stress induce actin accumulation, but the amount of actin accumulation induced by contractile activation is smaller (11).

These findings raise the question of the extent to which contractile activation (tone) modulates stress-induced changes in cell function and cytoskeletal remodeling. Here, we investigated this question in the cultured ASM cell by testing the hypothesis that reduction of ASM tone by broncodilators would inhibit stress-induced cytoskeletal stiffening and remodeling. We show that the greater the contractile tone the greater is the stress-induced cell stiffening and actin remodeling and that these changes are largely irreversible. As regards pathobiology of asthma, this suggests a maladaptive positive feedback in which tone potentiates ASM remodeling and stiffening that further increases stress and possibly leads to worsening airway function.

MATERIALS AND METHODS

Cell Culture

Canine ASM cells in culture were prepared as follows. Trachealis muscle was harvested and digested in collagenase and elastase with soy trypsin inhibitor as previously described (34, 36). Freshly dissociated cells were seeded into flasks at a density of 5 × 104 cells/cm2 in 1:1 DMEM and Ham’s F-12 (Invitrogen, Burlington, ON) with 10% fetal bovine serum, 100 U/ml of penicillin, and 100 μg/ml streptomycin. Medium was changed every 2–4 days, and the cells were passaged at confluence (10–14 days) before use. We used cells of passages 3–5 for experiments. Cells were serum deprived and supplemented with 5.7 μg/ml insulin and 5 μg/ml transferrin for 24 h before experiment. The cells were harvested by brief exposure to 0.05% trypsin and 1 mM EDTA, then plated onto either plastic 96 wells (6.4-mm diameter, COSTAR Stripwell, Corning, NY) at ∼20,000 cells/well or glass microscopic coverslips (12-mm diameter, Fisher Scientific, Nepean, ON) at ∼400,000 cells per coverslip. Preparations in 96 wells were used for measuring cell stiffness, and preparations on 12-mm coverslips were used for imaging and analyzing CSK structures subsequent to exposure to mechanical stress. Before cell plating, the plastic wells and coverslips were coated with type I collagen (Cohesion Technologies, Palo Alto, CA) at 25 and 5 μg/ml, respectively, for at least 24 h at 4°C.

Application of Stress to the Cell

Localized oscillatory mechanical stress was applied to the cell to induce CSK remodeling. This was done by oscillating magnetic microbeads that were tightly bound to the actin CSK using a device known as magnetic twisting stimulator (MTS). Details of this technique have been published elsewhere previously (11). Briefly, ferrimagnetic beads (4.5-μm diameter, produced by the laboratory of Dr. J. J. Fredberg of Harvard School of Public Health, Boston, MA) were coated with a synthetic Arg-Gly-Asp (RGD)-containing peptide at 50 μg peptide/mg beads (Peptite 2000, Integra Life Sciences, San Diego, CA) and stored in carbonate buffer (pH 9.4) at 4°C. Before use, beads were washed with phosphate-buffered saline (PBS) and resuspended in DMEM/F-12 medium with 1% BSA. Then, beads were deposited onto the surface of the cells prepared either in 96 wells or on 12-mm coverslips at ∼20,000 beads/well and ∼70,000 beads/coverslip, respectively. The RGD coating enabled the beads to bind specifically to transmembrane integrin receptors, forming focal adhesions, whereas the BSA blocked nonspecific binding. The binding between the bead and cell established in 15 min of contact and unbound beads were washed away subsequently.

Cells with surface-bound beads prepared as described above were placed inside the MTS, and the beads were first magnetized (∼2 kV, 250 μs) to align their magnetic moments horizontally each in the same direction. Then, a sinusoidal varying vertical magnetic field was applied, generating an oscillating torque to twist the magnetized beads. As a bead was tethered to the actin CSK through integrins and focal adhesions, the twisting motion of the bead thus applied a localized stress to the CSK, to which the bead was linked. The amount of magnetic torque applied on each bead was approximately the same and only depended weakly on the angle of rotation of the bead. However, the amount of stress applied to each cell varied from cell to cell due to the differences in the geometry of the bead-cell contact complex, as well as the number of integrin receptors within the contact complex. For a bead embedded 25% into the cell surface and a specific torque (torque normalized to bead volume) of 60 Pa, the stress applied to the cell was computed to be on the order of 1,000 Pa (27). In this study, we applied a specific torque of 56 Pa to the bead at a frequency of 0.3 Hz. This frequency was roughly equal to that of breathing. The MTS was placed inside an incubator (Sanyo). Thus, during mechanical stimulation, the cell was maintained at 37°C in humidified air containing 5% CO2.

Measurement of Cell Stiffness

To assess cell function during mechanical stimulation, we measured cell stiffness using optical magnetic twisting cytometry (OMTC). OMTC was similar to MTS in that it used magnetic beads in the same manner but in a smaller apparatus fitted on a microscope stage so that the bead motions can be optically tracked to determine the mechanical properties of the cell.

The detail of OMTC has been published elsewhere (13, 32). In brief, a 96 well containing cells with surface-bound beads, prepared as described above, was placed in a twisting device fitted on the stage of an inverted microscope (DM-IRB, Leica Microsystems). The microscope was equipped with a ×20 objective and a charge-coupled device camera (1,280 × 1,024 pixels, 12-bit gray scale, SensiCam, The Cooke, Auburn Hills, MI). Although the beads were twisted similarly as described above with a specific torque of 56 Pa at 0.5 Hz, the camera imaged the beads continuously (∼200 beads at a time) at 16 frames per twisting cycle. Then, using an intensity centroid algorithm, the bead positions in each recorded image were automatically determined (13). Then, Fourier transformation was used to extract the displacement of each bead in response to the applied torque, and beads with erratic, irreproducible motions were not analyzed (13, 27, 32).

Thus, for a given specific torque (T̃) applied to a bead and the resultant bead displacement (D̃), as described above, a complex stiffness (G̃) of the cell was defined as the ratio of the torque to the displacement, i.e., G̃ = (T̃/D̃) = G′ + iG”, where G′ is the in-phase component or elastic stiffness, which we hereafter refer to as cell stiffness (in Pascal/nm), G” is the out-of-phase component, and i is the unit imaginary number Math. In this study, we did not examine G", which is usually 10–20% of G′ (13).

Assessment of Actin CSK Remodeling

To assess actin CSK remodeling in response to stress, we prepared cells with surface-bound beads on 12-mm coverslips as described above. These cells were either stimulated by MTS or left unperturbed in an identical condition as time-matched control. At chosen time points during the stimulation, cells were washed in PBS and fixed in 4% paraformaldehyde in PBS for 15 min. Then the cells were thoroughly rinsed in PBS again and permeabilized with 0.3% Triton-X in PBS for 5 min. Once permeabilized, the cells were first incubated in PBS with 10% BSA for 1 h to block nonspecific staining of fluorescence and then incubated in PBS with ∼6.6 μM fluorescently conjugated phalloidin (Alexa 488, Molecular Probes, Eugene, OR) for 30 min to fluorescently label the filamentous actin (F-actin). Fluorescently labeled cells were rinsed in PBS and mounted on a glass slide with mounting medium (Prolong Antifade kit, Molecular Probes).

Using a confocal microscope (Leica TCS NT confocal scanning laser microscope, Heidelberg, Germany), a microscopy specialist who was unaware of the specific treatment of the specimen examined the actin CSK of the cell at ×100 magnification. Each coverslip was scanned in a cross pattern (x and y directions) from edge to edge. Along each path of scanning and at roughly equal separation, 10 fields of view were chosen for assessment of CSK remodeling. At each field, beads that were bound to cell surface were identified. By optically sectioning through the cell from the basal to the apical plane, the actin CSK in the neighborhood of each of these beads was thoroughly examined and imaged.

Subsequently, each image was examined by another person. Actin CSK in the neighborhood of each singular bead on the cell surface was analyzed and compared with the overall background of the surrounding region. Detail of this analysis to quantify bead-associated actin CSK remodeling has been published previously (11). Briefly, beads exhibiting discernable actin staining above background (noted by structure and/or intensity) in the neighborhood of the bead were scored as positive beads, and beads not exhibiting such actin staining were scored as negative beads. We thus quantified actin CSK remodeling as a percentage of positive beads in the total number of analyzed beads for each experimental condition. This gave a measure of the extent to which actin accumulation occurred in the neighborhood of the bead for a given bead population at a given time. We also analyzed the actin structures associated with the positive beads and characterized them as either small, hairline thin, and ring-shaped actin structures or larger, more intense actin structures as described previously (11). Thus a greater percentage of the positive beads that were associated with the larger actin structures indicates a greater amount of actin accumulation in the neighborhood of those beads.

Tone Modulation

Tone of the ASM cell was reduced by muscle relaxant agonist. A change in tone after administration of relaxant agonist was reflected by a change in cell stiffness measured by OMTC. This was based on several evidences that, over a wide range of contraction and relaxation, the intracellular contractile force as measured by contraction force microscopy is highly linearly associated with the cell stiffness as measured by OMTC (39, 42). Three different relaxant agonists, namely, dBcAMP, forskolin, and formoterol, were used at concentrations of 10−3, 10−4, and 10−6 M, respectively. These doses were the lowest that induced maximal changes in cell stiffness (4, 22), and the reduction in cell stiffness we observed here with these relaxant agonists was in good agreement with previously reported results using this technique (13, 20, 22, 25, 32).

Experimental Protocols

Measuring cell stiffness in response to application and cessation of stress.

Baseline cell stiffness (G′0) was measured in each well of cells prepared as described above. Subsequently, cells were placed either inside the MTS and continuously stimulated with stress or outside the MTS but in the same incubator as control. At 60 and 120 min, G′ was measured for both the stress-stimulated cell and the control. Then stress was ceased. The cells were left unperturbed for another 60 min in the incubator, and G′ was measured again (rest).

Measuring G′ in response to stress followed by relaxant agonist.

G′0 was measured in four wells of cells prepared as described above. Subsequently, cells in two of the wells were continuously stimulated with stress by MTS, and cells in the other two wells were placed in the same incubator as controls. At 60 min, G′ was measured for both the stress-stimulated cell and the controls. Afterward, 10 μl of 10−2 M dBcAMP were administered to one of the two wells with stress-stimulated cells, and 10 μl of 10−5 M formoterol were administered to the other (1:10 dilution). The same relaxant agonists were also administered to the controls. After administration of agonist, the cells were incubated for 5 min so that G′ reached a plateau as determined previously (13, 25). Then G′ was measured again to assess the agonist response. The experiment was repeated three to five times. Because each measurement of G′ took ∼30 s, the order for repeated measurement of G′ in different wells was randomized.

Measuring cell stiffness in response to stress in the presence of relaxant agonist.

G′0 was measured in four wells of cells prepared as described above. Subsequently, 10 μl of 10−2 M dBcAMP were then administered to each of the four wells (1:10 dilution for the 10−3 M dose), and all wells were incubated for 5 min. G′ was measured again in each well (time = 0 min). Then, cells in two wells were continuously stimulated with stress by the MTS, and the cells in the other two wells were not stimulated. At 5, 15, 30, 60, and 120 min, G′ was measured for both the stress-stimulated cells and the nonstimulated cells but all in the presence of the relaxant agonist. The experiment was repeated three to five times, and the order for measurement of G′ was randomized as described above. The same protocol was repeated on different cells with forskolin and formoterol.

Measuring actin CSK remodeling in response to stress in the presence of relaxant agonist.

Cells with surface-bound beads were prepared on 12-mm coverslips as described above. Three coverslips were placed in one 35-mm Petri dish containing 2 ml of medium, and four dishes were prepared for each experiment. Then, 200 μl of either agonist solution or media (controls) were administered to each dish, resulting in incubation of the cells in 10−3 M dBcAMP, 10−4 M forskolin, 10−6 M formoterol, or medium, respectively, for 5 min. Cells on coverslips, one from each of the four conditions, were fixed and stained for F-actin as described in Assessment of Actin CSK Remodeling (time = 0 min). Subsequently, cells on the remaining coverslips were continuously stimulated with stress by the MTS and fixed and stained as described above at time = 30 and 60 min. The experiment was repeated four times with randomized order of agonist administration and cell fixation.

Data Processing and Statistics

Because the G′ measured by OMTC was approximately log-normally distributed, the stiffness of each well was represented by the median stiffness of the measured bead population, and tests for significance of inter-well difference were performed on logarithmically transformed data. More than 200 beads were measured in each well. For groups of wells in repeated experiments with the same condition, the stiffness of each group was presented as means ± SE, where the mean is the G′ averaged over all wells in the group (n = 3–5). Due to variability in G′0, G′ was normalized to G′0 in each experiment so that the stiffness response to stress and/or agonist could be compared.

The percentage of actin-positive beads from image analysis was presented as means ± SE, where the mean was the average percentage and SE was from repeated experiments (n = 4). For each experimental condition, ∼100 beads on average (ranging from 51 to192) were analyzed. To compare two populations of analyzed beads from two different experimental conditions, a simple Student’s t-test with a 95% confidence level was used (P < 0.05) to test for significant differences. To compare multiple populations of analyzed beads from different experimental conditions, ANOVA for a single factor was used, instead, also with a 95% confidence level (P < 0.05).

Reagents

All chemicals were from Sigma Chemical (St. Louis, MO). Tissue culture reagents, including DMEM/F-12 medium and trypsin-EDTA solution, were purchased from GIBCO (Grand Island, NY). DBcAMP and forskolin were dissolved in ultra-pure filtered and sterile water at 10−2 and 10−3 M, respectively, as stock aliquots and stored at −20°C. Before use, these aliquots were quickly thawed and kept on ice. Formoterol was dissolved in DMSO at 10−2 M as stock aliquots and stored at room temperature. Before use, the stock aliquots of formoterol were further dissolved into ultra-pure filtered and sterile water at 10−5 M and kept on ice.

RESULTS

G′ vs. Application and Cessation of Stress

When the ASM cells were continuously stimulated with mechanical stress, G′ of the cells increased progressively to more than twofold of baseline (G′0) at 120 min (P < 0.05). By contrast, G′ of the controls increased only 56% from baseline in the same time. With rest (as described in the protocol) following cessation of stress, G′ of the stress-stimulated cells diminished only 24 ± 6%, leaving a substantial residual effect that was largely irreversible, and the stress-stiffened cells remained much stiffer than the controls (Fig. 1).

Fig. 1.

Change in cell stiffness (G′) in response to application and cessation of stress. Solid bars represent results for G′ of the stress-stimulated cell, and open bars represent results for G′ of the unperturbed control (control). G′ was measured at time = 0 [baseline stress (G′0)], 60, and 120 min in all cases. Then stress was ceased, both the stress-stimulated and control cells were left unperturbed for 60 min (rest), and G′ was measured again. For this and the other figures, unless noted, G′ is normalized to baseline stiffness (G′0). Values are means ± SE (n = 3–5), and P < 0.05 denotes a significant difference between groups (Student′s t-test).

Cell Stiffness vs. Stress Followed by Relaxant or Contractile Agonists

G′ of both the stress-stimulated cells and controls increased from baseline similarly as shown above. With the cells that had been stimulated with stress for 60 min, administration of 10−3 M dBcAMP and 10−6 M formoterol caused G′ to decrease by 27% (1.00 ± 0.03 to 0.74 ± 0.06 Pa/nm) and 36% (1.25 ± 0.15 to 0.79 ± 0.15 Pa/nm), respectively. Both dBcAMP and formoterol reduced the G′ of the stress-stimulated cell to a level not significantly different from that of the controls (P > 0.05, ANOVA), despite formoterol appearing to cause a greater effect (Fig. 2). Moreover, relaxant agonists appeared to have less effect on the stress-stimulated cell than on the nonstimulated control where the agonists reduced G′ by ∼33 and 46%, respectively (see minimum values in Fig. 3). We also measured G′ in response to contractile agonist (80 mM KCl) after 60 min of stress stimulation. In the stress-stimulated cells, KCl caused G′ to increase by 47.3 ± 8.6%, whereas in the controls KCl caused G′ to increase by only 21.5 ± 2.3%, indicating that the stress-stimulated cell was more responsive to contractile activation (P < 0.05).

Fig. 2.

Change in G′ in response to stress followed by relaxant agonist. Solid bars represent results for G′ of the stress-stimulated cell, and open bars represent results for G′ of the unperturbed control. G′ was measured at time = 0 min (G′0) and 60 min of stimulation with stress. Then the cell was incubated, unperturbed, for 5 min in relaxant agonist, with either dBcAMP (10−3 M) or formoterol (10−6 M check). G′ was then measured again (posttreatment). N.S. denotes no significant difference among the groups (ANOVA).

Fig. 3.

Change in G′ in response to stress in the presence of relaxant agonist. ▴, Results for G′ of the stress-stimulated cell; ○, G′ of the unperturbed control. Top: G′ was suppressed by dBcAMP initially, but gradually recovered to the baseline level or above at 60 min, regardless of stress. However, at time of <60 min, there was no difference in the G′/G′0 between the stress-stimulated cell and the control in the presence of dBcAMP, indicating that at these times, the stress-induced stiffening was completely abolished (P > 0.05). At time = 120 min, the stress-stimulated cell became stiffer compared with the control (*P < 0.05). Middle and bottom: forskolin (middle) and formoterol (bottom) suppressed G′ and inhibited the stress-induced stiffening similarly as dBcAMP, but with different efficacy.

Cell Stiffness vs. Stress in the Presence of Relaxant Agonist

When the cells were incubated in 10−3 M dBcAMP for 5 min before application of stress, G′ decreased by 28% from baseline (G′0 at time = −5 min). Subsequently, with the onset of stress at 0 min and continuous stimulation with stress in the presence of dBcAMP, G′ decreased further to 59% of G′0 at 5 min. Afterward, G′ appeared to increase progressively in all cases. However, the normalized stiffness (G′/G′0) of the stress-stimulated cells showed no difference compared with that of the unperturbed controls until 60 min, indicating that at such times the stress-induced stiffening was completely abolished by the presence of dBcAMP. At 120 min, G′/G′0 of the stimulated cells became greater than that of the controls (P < 0.05) (Fig. 3, top).

G′/G′0 changed similarly in response to stress in the presence of forskolin and formoterol (Fig. 3, middle and bottom). However, forskolin (10−4 M) reduced G′ to a great extent (86% decrease from G′0 at 0 min) but lasted for a short duration (<15 min). After 15 min, the stress-stimulated cells gradually stiffened more than the controls, but the differences in G′/G′0 between the two conditions remained <30% for time of <60 min and was 44% at 120 min. Thus the presence of forskolin resulted in a partial but substantial inhibition of the stress-induced cell stiffening. The long-acting β-agonist formoterol (10−6 M) was able to suppress G′ below G′0 at all times during the 2-h experiment. Even more impressively, the presence of formoterol completely prevented the stress-induced cell stiffening during the 2-h experiment as the stress-stimulated cells showed no difference in the G′/G′0 compared with that of the controls at all times (P > 0.05).

The inhibitory effects of relaxant agonists on stress-induced cell stiffening are further illustrated in Fig. 4, where the difference in G′/G′0 between the stress-stimulated cells and the controls [i.e., G′/G′0(stress) − G′/G′0(control) %] in the absence or presence of relaxant agonist was plotted against time. In the absence of relaxant agonist, stress induced a time-dependent increase in cell stiffness, leading to a 122% greater increase of G′/G′0 in the stress-stimulated cells than in the controls. In contrast, the presence of a relaxant agonist largely inhibited the stress-induced stiffening, resulting in no more than a 35% increase of G′/G′0 in the stress-stimulated cells compared with the controls, regardless of the agonist used.

Fig. 4.

Stress-induced cell stiffening in the absence or presence of each of the relaxant agonists. The extent of stress-induced cell stiffening is given as a %increase in G′ that was computed as the difference in G′/G′0 (%) between the stress-stimulated cell and the control. It shows that the stress-induced stiffening was greatly reduced in the presence of a relaxant agonist (P < 0.05; time ≥ 30 min).

In the presence of a given relaxant agonist, the suppression of G′ from G′0 indicates the agonist’s ability to ablate tone, and the duration within which the G′/G′0 of the stress-stimulated cells was not statistically different from that of the controls indicates the agonist’s ability to inhibit stress-induced cell stiffening. When these two factors were both compared among the three relaxant agonists used, formoterol emerged as the most effective, both ablating tone and inhibiting stress-induced cell stiffening throughout the 2-h mechanical stimulation. Perhaps not surprisingly, this agrees with the fact that the long-acting β-agonist remains effective in bronchodilation for extended periods, although the mechanisms are not understood (3). Incidentally, this suggests that such in vitro measurement with a relaxant agonist or bronchodilator agent might be useful to evaluate its efficacy in mitigating stress-induced changes in the ASM cell.

Actin Structures vs. Stress in the Presence of Relaxant Agonist

In all cases, whether the cell was stimulated with stress in the absence or presence of any of the relaxant agonists, the percentage of beads with positive actin staining increased similarly to that reported previously (11), reaching ∼60% at 60 min (Fig. 5). There appeared to be no differences in the percentage of actin-positive beads among the stress-stimulated cell groups at any time, regardless of the presence of relaxant agonist (ANOVA, P > 0.05). However, relaxant agonist did appear to inhibit the growth of actin accumulation in the neighborhood of the bead during stress application. This is shown in Fig. 6, where, in addition to the percentage of actin- positive beads, the number of positive beads associated with either large or small ring-shaped actin structures in the stress-stimulated cells in the absence of relaxant agonist was compared with that in the presence of formoterol. In the cells stimulated with stress alone, more beads exhibited large and extensive structures such as the spindle-like structures shown in Fig. 6A, and as previously reported (11). In contrast, in the cells stimulated with stress in the presence of a relaxant agonist, most positive beads were associated with hairline ring structures with no apparent extensions away from the bead, as shown in Fig. 6B. At 60 min after stress application, 62.9 ± 5.1% of positive beads were associated with large actin structures in the cells stimulated with stress alone vs. 27.8 ± 4.1% in the cells that were stimulated with stress in the presence of formoterol (P < 0.05) (Fig. 6, left).

Fig. 5.

Actin cytoskeleton (CSK) remodeling, quantified as the percentage of beads associated with positive fluorescent actin staining of the total number of beads measured vs. time. Solid bars with or without labels represent results from the stress-stimulated cells with or without relaxant agonist. Open bars are results of the time-matched control. Values are means ± SE (n = 4). *Significant difference from time-matched stress-stimulated cells (P < 0.05).

Fig. 6.

Comparison of actin structures formed in the neighborhood of beads during stimulation with stress in the absence or presence of relaxant agonist. In the absence of relaxant agonist, many positive beads from the stress-stimulated cells at 60 min exhibited large extensive actin structures such as the spindle-like structure shown in A. In contrast, in the presence of relaxant agonist, the positive beads exhibited predominantly small hairline ring structures, such as the one shown in B. Left: percentage of beads associated with either large or ring structures among the positive beads from the cell either stimulated with stress alone for 60 min or stimulated with stress for 60 min in the presence of formoterol. Evidently, a greater portion of the positive beads in the cells stimulated with stress alone were associated with large actin structures compared with those in the cells stimulated with stress in the presence of formoterol (P < 0.05).

DISCUSSION

The primary findings of this report are as follows. In cultured ASM cells, continuous application of localized oscillatory mechanical stress induced time-dependent cell stiffening and localized actin remodeling. These changes were modulated by the level of contractile tone. In the presence of baseline tone, which was appreciable, mechanical stress could induce a twofold increase in cell stiffness, and this cell stiffening largely persisted following stress cessation alone. This stiffening could be prevented, however, by ablation of tone before application of mechanical stress. Ablation of tone, although it had little effect on the initiation of actin formation at the sites of stress, did inhibit the total amount of actin accumulation. These observations thus indicate that the greater the contractile tone, the greater was the stress-induced cell stiffening and localized actin remodeling. In the remainder of this discussion, we deal with methodological limitations, discuss the results, and then conclude by describing why these results might be important in asthma.

We used magnetic beads that were tightly attached to the CSK, both to measure cell stiffness, as has been used previously (9, 11, 25), and to apply localized cyclic mechanical stress to induce actin remodeling. It has been established previously that the stiffness measured by this method corresponds well with that measured by other methods, such as atomic force microscopy (1, 13, 29), and reflects mechanical properties of deep CSK structures as verified by intracellular stress tomography (19). Regarding cell stiffness, limitations of the bead-twisting technique have been discussed in detail elsewhere (12, 25, 27, 29, 32). Regarding CSK remodeling, the following technical limitations of the methods used here are of concern. Primary among these concerns was heterogeneity from cell to cell in the stresses applied by the bead. Heterogeneity in the locally induced stress could be attributable only to differences from cell to cell in geometry of the bead-cell complex (27) and intrinsic cell stiffness (13). Here, we studied a sufficiently large number of beads that the standard error of measures of cell stiffness and remodeling, respectively, were acceptably small. Another important limitation was that, in a given bead, we could not simultaneously follow the progression of actin accumulation and stiffness changes. Instead, we measured actin remodeling in cells fixed and stained after the experiment, and these measurements were in good agreement with our laboratory’s earlier findings (11).

In the presence of baseline tone, the stress-induced stiffening and remodeling in the ASM cell reported here were consistent with previous reports on stress-stimulated ASM cells (11, 3237) and was also in general agreement with reports in other cell types (15, 24). In the fibroblast and the endothelial cell, for example, a localized stress applied by a microbead induces CSK remodeling, including accumulation of actin filaments and formation of focal adhesion complexes (15, 24). Such remodeling leads to strengthening and stiffening of local structures (9, 15). It has also been demonstrated in these cell types that mechanically induced CSK remodeling and stiffening are regulated by intracellular contractile forces (5, 10, 15).

In the ASM cell, we show here that the cell stiffening and localized actin remodeling in response to localized stresses were dependent on the preexisting level of contractile activation (tone) (Figs. 2, 3, and 6). At baseline tone, continuous stimulation with localized stresses led to a great extent of cell stiffening, and the stress-induced cell stiffening persisted for >1 h after cessation of the mechanical stimulation but could be reduced when tone was diminished (Figs. 1 and 2). More importantly, when tone was reduced from baseline before imposition of stress, the stress-induced cell stiffening observed in the presence of baseline tone could be largely inhibited or even completely abolished (Fig. 3). As regards actin remodeling in response to continuous stimulation with stress, tone ablation from the baseline led to much less actin accumulation at the sites of stress application (Fig. 6).

Mechanisms through which ASM tone regulates stress-induced cell stiffening and actin remodeling are not clear. In other cell types, it has been shown that mechanotransduction of external forces into discrete signaling cascades depends on the mechanical properties of the focal adhesion and linked actin stress fibers (7, 15). The formation of these structures depends on the forces exerted through them, whether generated internally via contractile activation or externally by applied stress, and the structures disintegrate on cell relaxation (7, 15). There are several pathways known to mediate stress-induced formation of focal adhesions and actin assembly, recently reviewed by Bershadsky et al. (7). These include conformational changes at focal adhesions in response to applied mechanical stress that can alter integrin affinity for focal adhesion kinase, triggering focal adhesion growth and actin accumulation. Actin polymerization can also be promoted through the opening of stretch-activated calcium channels, as well as stretch-induced activation of a variety of actin assembly proteins such as Rho-associated kinase, p38 mitogen-activated protein kinase, and small heat shock proteins (HSPs) such as HSP27. Interestingly, it was recently shown in fibroblasts that force-induced actin assembly predominantly involved smooth muscle α-actin rather than β-actin and that the stress-induced activation of p38 depends on the polymerization of smooth muscle α-actin, which may mean that the organization of contractile filament networks is more sensitive to applied mechanical stress than that of cytoskeletal β-actin networks (41). However, the influence of contractile activation on actin remodeling has been largely studied in conditions where no external stress is applied. The studies presented here establish that stress-induced actin polymerization depends on the level of cell activation.

With regard to contractile activation, it is well established that contractile activation leads to actin polymerization. For example, in intact tissue strips, Mehta and Gunst (26) demonstrated that activation by acetylcholine stimulates actin polymerization and directly contributes to force development that is thought to be independent of regulatory pathways for myosin light chain phosphorylation. They suggested that contractile activation initiates mechanically sensitive signaling pathways promoting actin polymerization. Indeed, in the cultured ASM cell, a variety of contractile agonists have been shown to induce actin polymerization by activating pathways that are also activated by applied mechanical stress, such as those involving the Rho family of proteins (2, 13, 17, 20, 25). Therefore, contractile activation and mechanical stress could induce remodeling and stiffening through common pathways. If so, inhibition of tone would inhibit mechanically transduced intracellular events such as the actin remodeling that we observed in the presence of baseline tone. However, few have directly studied the effect of cell relaxation on actin remodeling in the ASM cell. Hirshman and coworkers (18) reported that relaxation of the ASM cell by relaxant agonist such as forskolin or isopreterenol leads to actin depolymerization, consistent with the idea that tone inhibition would decrease actin polymerization due to applied mechanical stress.

In the present study, we compared the effects of three different relaxant agonists on the duration of tone depression and the ability to inhibit cell stiffening and actin remodeling. Although each reduces contractile tone, each achieves its effect through slightly different routes with different effective duration. dBcAMP increases intracellular cAMP directly, whereas forskolin activates the enzyme adenylyl cyclase to catalyze the conversion of intracellular ATP into cAMP. Formoterol, on the other hand, binds to the β-adrenoceptors on the cell membrane, resulting in the release of the G protein’s α-subunit that activates adenylyl cyclase. By any route, an increase in cAMP activates protein kinase A, which in turn phosphorylates myosin light chain kinase, leading to increased actin-myosin cross-bridge detachment and ASM relaxation. However, increased cAMP also promotes relaxation via different mechanisms, including K+ channel alteration, Na+-K+-ATPases, Ca2+ influx, Ca2+ sequestration, Ca2+ sensitivity on myosin, and IP3 formation (16, 40).

With these three different relaxant agonists, the results show that forskolin, which depressed tone briefly, inhibited stress-induced stiffening for only 15 min, whereas dBcAMP and formoterol, which depressed tone longer, inhibited stress-induced stiffening for 60 min and over 2 h, respectively (from Figs. 3 and 4). Our results also show that direct manipulation of cAMP or indirect manipulation by stimulation via β-adrenergic receptors produced similar inhibition of actin remodeling during the period when tone was depressed.

In the presence of baseline tone, actin accumulated in the vicinity of the beads, resulting in the appearance of small ring-shaped actin structures. Such structures have been observed in different cell types, including the ASM cell (11, 15, 19, 29). Actin accumulates even further with continuous application of oscillatory stresses, leading to larger and more intense actin structures close to the bead (11). When the cell is activated above baseline tone by exposure to contractile agonists, the number of beads exhibiting small ring-shaped actin structures increases (11). Here, we show that when the cell was relaxed below baseline tone by exposure to relaxing agonists, the number of beads exhibiting small ring-shaped actin structures did not change (Fig. 5). In the relaxed cell, oscillatory stress failed to induce further actin accumulation (Fig. 6). Taken together, these findings show that actin accumulates in the form of small ring-shaped structures, regardless of the level of contractile tone and regardless of the presence of imposed stress. However, further actin accumulation in the form of larger structures occurs only in the presence of both imposed stress and appreciable tone.

We speculate that the formation of the small ring-shaped actin structures might be initiated in the vicinity of a bead due to the formation of small focal complexes. These focal complexes are mechanical sensors first described in lamellipodial protrusions, and their formation does not depend on the level of contractile activation (7). In the presence of tone, because of the resistance of internal contractile forces, the mechanical signal due to oscillatory bead motion might be transmitted in distance to induce further actin accumulation, increasing the size of the actin structures. This would lead to strengthening of local CSK and increased cell stiffness. In the absence of tone, however, due to lack of internal resistance, the imposed stress might induce large deformation of the cell but not distant mechanotransduction. Thus further actin accumulation and cell stiffening during stress stimulation could be inhibited.

Modulation of cytoskeletal stiffness and structure in the ASM by tone might have important implications in asthma. We suggest that tone might play dual but competing roles in regulating the way in which mechanical stress affects ASM function in asthma. On one hand, stresses associated with tidal breathing and deep inspiration (DI) are beneficial for ASM function in healthy individuals. In normal individuals, a DI is known to be the most potent of all known bronchodilator agencies and can reverse or prevent spasmogen-induced airway constriction (8, 21, 30). Cyclical stretch of ASM reduces force generation and stiffness or, depending on the loading conditions, causes muscle lengthening; these phenomena have been suggested as a putative mechanism for the dilatory effects of DI (14). As such, imposed stresses relax the cell and thus further reduce ASM tone. On the other hand, tidal stresses associated with breathing may cause CSK reorganization, leading to enhanced contractility (11, 32, 3437), and here we have shown that tone enhances ASM contractile function and remodeling. Because asthma is associated with greatly enhanced ASM tone (28), this implies that the pathogenesis of ASM in asthma might be in part caused by the imposed stresses. This suggestion is consistent with the observation that the dilatory effects of DI are somehow impaired or even reversed in asthma (30) and that the ASM cell obtained from patients of asthma exhibits increased proliferation, increased shortening velocity and capacity, and increased myosin light chain kinase content (6, 23, 43) as observed in the stress-stimulated cell in culture (34, 36, 38). This raises the possibility of a maladaptive positive feedback in asthma in which tone potentiates stress-induced ASM remodeling and stiffening that, in turn, further increases stress, thus forming a vicious cycle of worsening airway function (Fig. 7). If so, then continuous reduction of bronchomotor tone via treatment with long-acting β-agonists such as formoterol may break this cycle.

Fig. 7.

Asthma and bronchial hyperresponsiveness (BHR) are associated with elevated bronchial tone in airway smooth muscle (ASM). This may potentiate the ability for mechanical stress to cause deleterious effects on ASM, such as cell proliferation and CSK remodeling. These changes may lead to an increase of ASM stiffness and contractility that, in turn, further increases mechanical stress, eventually forming a vicious cycle. Airway inflammation and airway remodeling associated with asthma may also increase stress, amplifying the cycle. If present in vivo, the cycle described above may provide a new mechanism for worsening airway function. If true, then reducing tone may act to break this cycle, and this might partly explain why using the combination of corticosteroids with long-acting β-agonists that act to reduce ASM tone is more effective than using corticosteroids alone (31).

GRANTS

This study was supported by a Whitaker Foundation Research Grant, National Heart, Lung, and Blood Institute Grant HL-65960, and the Canadian Foundation for Innovation, the Nova Scotia Health Research Foundation, and a Legacy Grant of Nova Scotia Lung Association.

Acknowledgments

We thank Jean Lai of Harvard School of Public Health for the fluorescence confocal microscopy imaging. We also thank Drs. Fabry, Shore, and Trepat for valuable suggestions regarding using OMTC, β-adrenoceptor agonists, and data interpretation.

Footnotes

  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

REFERENCES

View Abstract