INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

MANY STUDIES HAVE SHOWN THAT deep inspiration (DI) comprises the first line of defense against bronchospasm in normal subjects (5, 11). It has also been suggested that an impairment of the ability of inspiration to stretch contracted airway smooth muscle (ASM) is an important feature of altered airway function in asthma (2, 11, 12, 17), although these studies do not exclude the possibility that a myogenic response from asthmatic ASM could play a role in resisting the relaxing effect of DI. Malmberg et al. (10) were the first to show that taking a DI before administration of a bronchoconstricting stimulus attenuated the subsequent response in normal subjects. Since then, a number of investigators have confirmed that DI is bronchoprotective as well as bronchodilating (7, 14, 15, 17), and Chandy et al. (1) have shown that a fast DI is more effective than a slow DI. The mechanisms of DI-induced bronchodilation and bronchoprotection are unknown. Fredberg et al. (3, 5) have hypothesized that perturbed equilibrium of myosin binding is the basis for the bronchodilating effect; the mechanical strain produced by a DI or tidal breathing is transmitted to the myosin heads, causing them to detach from the actin filaments much sooner than they would under isometric conditions, and decreasing the number of attached cross bridges, thereby decreasing the active force.

Although cross-bridge dynamics can help to explain the bronchodilating effect of DI, they cannot explain the subsequently reduced bronchoconstriction when DI is elicited before stimulation (bronchoprotective effect) (14, 15). ASM plasticity could explain the bronchoprotective effect. Smooth muscle plasticity is characterized by the muscle's ability to adapt to length changes by reorganizing its contractile apparatus. To accommodate large changes in cell geometry without compromising its contractile function, smooth muscles might have evolved a mechanism that disassembles and reassembles the contractile and cytoskeletal filaments to achieve optimal filament overlap. Because structural reorganization takes time, monitoring the time course of tension change after a length perturbation could produce insights into the process of subcellular reorganization. The plasticity model predicts that there will be an initial decrease in isometric tension after a large length change, due to disassembling of contractile units before they are reassembled again to suit the new cell dimension. The model also predicts that there will be tension recovery after a length change, if the muscle is allowed to adapt to the new length. Some of these predictions have been observed (6, 13). The plasticity model, therefore, could explain the bronchoprotective effect of length perturbation stemming from DI and tidal breathing.

Admittedly, the exact mechanism of smooth muscle plasticity is not known. The present functional study is not designed to elucidate the underlying mechanism of plasticity. The results from this study, however, could yield valuable insights into the mechanism, which could then be used to guide studies with ultrastructural and biochemical approaches. An important focus of this study is to correlate the in vivo observations of the effects of DI and tidal breathing to in vitro behavior of ASM.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Tissue preparation. Segments of trachea were removed from pigs (body wt = 20-25 kg) immediately after the animals were killed. The segments were immersed in physiological saline solution at 4°C [composition (in mM): 118 NaCl, 5 KCl, 1.2 NaH₂PO₄, 22.5 NaHCO₃, 2 MgSO₄, 2 CaCl₂, and 11.1 glucose]. The solution was aerated with 95% O₂-5% CO₂ to maintain a pH of 7.4. Rectangular strips of trachealis muscle were dissected from the trachea after removal of the epithelial and connective tissue layer. Muscle strips (6-7 × 2-3 × 0.5 mm) were vertically mounted in a tissue bath, with one end fixed to a stationary hook and the other attached to the servo-controlled lever of a length-force transducer. The tissue bath contained solution at 37°C that was bubbled with 95% O₂-5% CO₂.

Length oscillation applied to muscle strips. Length oscillation was applied to the muscle through the servo-controlled lever system. Sine waves were used, and the amplitude, frequency, and duration were varied independently. Triangular waves of a single period and different frequencies were also used, in separate tests, to compare the effects of slow and quick stretches. Length oscillations were imposed on the muscle at L_ref. The selected stretch amplitudes of the oscillation were between 4 and 34% of L_ref; the selected frequencies were 0.25, 0.5, 0.75, and 1 Hz; and the durations were one cycle (to mimic a single DI), 20 s, and 1, 2.5, 5, and 10 min. When the amplitude was varied, the frequency was set at 0.5 Hz (assumed to be the approximate breathing frequency of pigs) and the duration was set at 5 min. When the frequency was varied, the amplitude was chosen to be 29% L_ref and the duration was again 5 min. To test the effect of the duration, the amplitude and frequency were set at 29% L_ref and 0.5 Hz, respectively. Single cycle triangular waves of two frequencies [0.25 Hz (slow) and 1 Hz (quick)] were produced with a stretch amplitude of 29% L_ref. All oscillations were performed when the muscle was at resting condition.

Correction for system compliance. The compliance of the lever system itself is negligible; however, the surgical silk that connected the muscle to the lever system had a small but significant compliance. The total system compliance (lever + silk) was obtained by applying calibrated forces to the system and the silk thread without a muscle strip and recording the length changes. The system compliance was then subtracted from the measured total compliance to obtain the correct length oscillation amplitude.

Data analysis. The decrease in isometric force during the stimulation immediately after the oscillation was used as a measure of force reduction and was expressed as a fraction of F_max. The subsequent isometric measurements were used to determine the time course of the force recovery. The amplitude of the oscillation was expressed as amplitude of stretch, which was one-half of the oscillation amplitude. The effect of the amplitude was analyzed using linear regression. Comparisons of the mean force reductions after oscillation at different frequencies and durations were made using linear regression and one-way ANOVA. Comparisons of force reduction after quick and slow triangular or ramp stretches were made using a paired t-test and a sign test. The time course of force recovery was fitted to the simple 2-parameter exponential equation [y = a(1exp^bx)], where a and b are constants. Comparison of the b values (rate of recovery) for different time intervals was made with one-way ANOVA. For all statistical tests used, P > 0.05 was considered statistically insignificant. In each case, n is the number of muscle strips used for each experiment.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Effect of amplitude. The fraction of F_max, calculated from the first contraction immediately after the oscillation, is plotted against the amplitude of stretch in Fig. 1. At the lowest chosen amplitude of stretch (4% L_ref), the active force was reduced to ~95% F_max. A significant linear correlation (r) was found between the force and the amplitude of stretch (r = 0.81, P < 0.0001). The extent of the force reduction was substantial. At a stretch amplitude of 29% L_ref, the oscillation reduces the subsequent force generation by 24.9 ± 1.9% (n = 9). Oscillation frequency used for this group of measurements is 0.5 Hz, and the duration of the oscillation is 5 min.

View larger version (13K): [in this window] [in a new window]   Fig. 1.   Effect of amplitude of stretch on subsequent active force development. , Active force measured from individual preparations immediately after the oscillation, expressed as fractions of maximal isometric force (Fmax). Oscillation frequency and duration are 0.5 Hz and 5 min, respectively. Linear regression shows significant correlation (r = 0.81, P < 0.001). Lref, reference length of sample.

Effect of oscillation duration. The means and SE of the fraction of F_max calculated at each chosen duration are presented in Fig. 2. The force reduction is evident even at durations as short as one cycle. A significant linear correlation was present (r = 0.98, P = 0.0029). One-way ANOVA showed that the difference among the mean values was significant (P = 0.002), i.e., the longer the duration, the greater the force reduction. A 10-min oscillation produces a force reduction of ~23.1 ± 2.0% (n = 6). Oscillation frequency used for this group of measurement was 0.5 Hz, and the amplitude of stretch was 29% L_ref.

View larger version (10K): [in this window] [in a new window]   Fig. 2.   Effect of duration in seconds of oscillation on subsequent active force development. , Mean values of active force measured immediately after the oscillation, expressed as fractions of Fmax. Oscillation frequency and stretch amplitude are 0.5 Hz and 29% Lref, respectively. Linear regression shows significant correlation (r = 0.98, P = 0.0029). Error bars represent SE.

Effect of oscillation frequency. The extent of force reduction in a 5-min-duration sine wave oscillation had no significant correlation to oscillation frequency, as shown in Fig. 3. A straight line was fitted to the mean values, but the P value of the slope was 0.184 (i.e., P > 0.05). One-way ANOVA also showed that the mean values for different frequencies were not different (P = 0.699). However, the rate of a single-cycle ramp stretch and release applied to the muscle (n = 6) affected the subsequent force reduction, as presented in Fig. 4. The difference in active force reduction between a 0.25-Hz and a 1-Hz triangular wave oscillation of one cycle was not significant (P = 0.12) according to the paired t-test but was significant (P = 0.03) according to the sign test (8). The amplitude of stretch used for this group of measurements was 29% L_ref, and the duration of oscillation was 5 min, excepting the one cycle oscillation.

View larger version (9K): [in this window] [in a new window]   Fig. 3.   Effect of oscillation frequency on subsequent active force generation. , Mean values of active force measured immediately after the oscillation, expressed as fractions of Fmax. Stretch amplitude and duration are 29% Lref and 5 min, respectively.

View larger version (12K): [in this window] [in a new window]   Fig. 4.   Effects of a slow (0.25 Hz) and a quick (1 Hz) single stretch on the subsequent active force generation. Active force was measured from individual preparations immediately after the oscillation and is expressed as fractions of Fmax. Amplitude of ramp stretch was 29% Lref. Each line connects two measurements made from the same preparation.

Isometric force recovery after oscillation. After the length oscillation, isometric force was reduced to a level below F_max; it then gradually recovered to the preoscillation level after a series of isometric contractions over a period of ~30 min. The interval between contractions had an effect on the rate of the force recovery. The measured isometric force from each contraction was converted to percentage of total recovery. They were then fitted to the exponential function y = a(1exp^bx), in which y was percent total recovery and x was time. The correlation values ranged between 0.999 and 0.9999. The means (n = 6) and SE of the calculated percent recovery, as well as the exponential fits, are shown in Fig. 5. The rates of recovery (b) for the 5-, 7.5-, and 10-min intervals were 0.24, 0.22, and 0.17 min¹, respectively. They were obtained using the same curve- fitting procedure and were significantly different (P = 0.027) according to one-way ANOVA.

View larger version (16K): [in this window] [in a new window]   Fig. 5.   Active force recovery after length oscillation. The active force values are expressed as percentage of total recovery. The means of 5 samples and SE of percent recovery are plotted vs. time. The exponential function fitted to the means is of the form y = a(1expbx), in which y is the percent recovery and x is time. Inset: values for the rate of recovery (b) vs. frequency of stimulation.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

It appears that DI have both bronchodilating and bronchoprotective effects. For the sake of clarity in the following discussion, the bronchodilating effect of DI is defined as the reduction in bronchoconstriction resulting from the act of DI after the ASM has been contracted by administration of stimulants. The bronchoprotective effect of DI, on the other hand, is defined as the reduction in bronchoconstriction resulting from the act of DI before the ASM has been stimulated to contract. In the present study, a reduction in isometric force is interpreted as an in vivo equivalent of reduction in bronchoconstriction. This assumption is valid if all other factors affecting the ability of ASM to contract remain constant. Since 1961 (12), it has been known that DI has a bronchodilating effect. The theory of perturbed myosin binding, recently articulated by Fredberg et al. (3), suggests that the bronchodilating effect of DI is related to disruption of cross-bridge dynamics. However, Malmberg et al. (10) and Shen et al. (15) reported that DI performed before stimulation also reduced the subsequent bronchoconstriction. This bronchoprotective effect of DI cannot be explained by disruption of cross-bridge dynamics because the length perturbations are applied before the cross bridges are activated.

Non-cross-bridge mechanisms play a role in regulating the effect of length oscillation on smooth muscle contractility (16). Evidence (6, 13) has shown that the organization of contractile filaments within smooth muscle cells can be modified in response to changes in muscle length and, therefore, can affect the contractile response of the muscle.

Our observation of a temporary loss of the ability of ASM to generate maximal force after a length perturbation could explain, at least in part, the in vivo observation of the bronchoprotective effect of DI. It is not clear what is responsible for this temporary force reduction. It is likely that the structure of the contractile apparatus could be disrupted by a length perturbation; however, the nature of the disruption is also not clear. It could involve a rearrangement in the contractile filament overlap, partial dissolution of the labile myosin thick filaments, or dislocation of the actin anchoring sites (i.e., dense bodies and dense plaques). When a length perturbation is applied to an activated muscle, the disruption of cross-bridge binding accounts for most of the observed force reduction (4). However, the length perturbation could also disrupt the structure of the whole contractile apparatus, as suggested by the present study. This could explain the finding of Fredberg et al. (3) that disruption of cross-bridge binding cannot account for all the observed force loss due to oscillation, as discussed in the following section.

Amplitude effect and force recovery. An amplitude effect similar to that depicted in Fig. 1 was noticed by Fredberg et al. (3). In their study, oscillations with increasing amplitude were imposed on an active, preshortened muscle; the oscillations caused lengthening in the muscle, suggesting that the bronchodilating effect of tidal breathing could be due to periodic length changes in the muscle. Their proposed model of perturbed cross-bridge binding explains part of their observations. However, the model was not able to account for their observation that, after a period of large-amplitude oscillation, the lengthened muscle was not able to immediately return to its original, shorter length on returning to a smaller amplitude of oscillation. Their observation indicates that there are factors other than cross-bridge interaction that determine the extent of muscle relaxation or force reduction after oscillation. Our observation suggests that large-amplitude oscillation causes a relatively long-term, although temporary, loss of force that cannot be explained by the perturbed cross-bridge binding mechanism. A more plausible explanation is that the perturbation results in a larger scale disruption that involves rearrangement of the contractile machinery in the muscle. It should be pointed out that, after the oscillation has been terminated, effects of the perturbation persist for a relatively long period of time. The time course of force recovery follows an exponential process, perhaps reflecting a recovery process that involves the contractile apparatus reorganization, including rearrangement of the contractile filaments, repolymerization of the myosin thick filaments, and reanchoring of the actin attachment sites. The observation that force depression is a linear function of amplitude of oscillation (Fig. 1) favors a model in which reorganization of the contractile filaments is initiated by a signal from some unidentified stretch receptor that has an output proportional to the amount of stretch. It should be pointed out that there are numerous receptors on a muscle cell that could respond (specifically or nonspecifically) to a mechanical stretch and produce effects on the cell that either enhance or reduce isometric force production.

Duration effect. The effects of oscillation duration on the subsequent force development are shown in Fig. 2. It appears that durations as short as 0.5 s (a single cycle) show an inhibitory effect on the subsequent force production. Longer durations (e.g., 5 and 10 min) produce a more pronounced depression in force. The mechanism underlying this observation is not known. If length oscillation induces structural reorganization in the contractile apparatus via stretch receptors, then, theoretically, a single cycle should be able to initiate the changes, and additional cycles with the same amplitude should not produce further inhibition in force as suggested by Fig. 1, which indicates that force reduction is a function of oscillation amplitude. The fact that a prolonged period of oscillation produces a greater depression in force suggests that force inhibition by length oscillation is not a simple function of oscillation amplitude but also depends on other unknown factors. These data suggest that length perturbation associated with tidal breathing in normal subjects can create a state in which the ASM generates only a fraction of force relative to its maximal potential.

Frequency effect. In the studies by Shen et al. (16), length oscillations were applied to an activated muscle and oscillation frequency was found to be correlated to force depression. In the present study, the oscillations are terminated before muscle activation, and no correlation between oscillation frequency and force reduction is found (Fig. 3). The results suggest that length oscillation affects active and passive muscles in different ways. In the previous study, cross-bridge binding is sensitive to the frequency of length release and stretch; in the present study, the frequency does not appear to be an important factor determining the extent of force reduction. It should be pointed out that, in this group of experiments, the frequency effect is examined after a prolonged period (5 min) of oscillation duration. In a separate group of experiments, single ramp stretch and release was applied to the muscle at different ramp speeds, and the results were different. A quicker stretch produces a slightly greater depression in the subsequent force generation (Fig. 4). Although the difference in force reduction is small, it is significant by sign test. The physiological significance of this difference is not clear, although it is interesting to note that in vivo tests on normal subjects indicated that a fast DI produced greater bronchodilation than that produced by a slow DI (1).

Rate of force recovery and stimulation frequency. Force recovery after a length perturbation is likely an active process, as suggested by the dependence of the rate of force recovery on the frequency of muscle activation (Fig. 5). It is not clear how activation facilitates force recovery; however, it is possible that phosphorylation and/or dephosphorylation of the contractile and cytoskeletal proteins may be required for structural reorganization. A similar force recovery is observed in tracheal smooth muscle subjected to a step length change (13); force reduction associated with the length change is fully recovered after six stimulations at 5-min intervals, and full recovery is obtained even when the muscle is subjected to a 300% length change. This ability of tracheal smooth muscle to adapt to length changes is explained by a theory, briefly outlined in our opening paragraphs, which postulates that the plastic reorganization occurring in smooth muscle cells enables the muscle to accommodate large length changes (13). A change in muscle length is likely the stimulus that signals the cell to undergo plastic rearrangement. The length change can be a step stretch or release, or it can be a length oscillation. The exponential time course of force recovery is likely a reflection of the process of subcellular reorganization, and this process is both time and activation dependent, as suggested by the present results.