INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

THE DEPENDENCE OF ISOMETRIC force on muscle length is thought to be governed mainly by the amount of overlap between myosin thick filaments and actin thin filaments. Although this is true for striated muscle, recent experiments have suggested that a different mechanism may underlie the length-tension (L-T) relationship in smooth muscle (3, 9). The ability of smooth muscle to adapt to length change confers a broad plateau on the L-T curve. Harris and Warshaw (4) found that, in single isolated smooth muscle cells, the L-T curve for the preparation shifts according to the initial length of the muscle, if the initial length is set when the muscle is at rest. The notion of optimal length (L_o) in such an L-T curve is meaningless because the length at which optimal tension (P_o) occurs will shift depending on the length at which the muscle is adapted. Shifting of the L-T curve (or L_o) in response to chronic passive shortening has been described in skeletal muscle such as the diaphragm in emphysematous hamsters (2, 7). The diaphragm muscle overcomes the disadvantage of the short working length because of emphysema by deleting some of the sarcomeres in series; this enables the thick and thin filaments in the muscle to achieve optimal overlap. The mechanism for such adaptation in skeletal muscle is probably different from that in smooth muscle in which rearrangement of the contractile filaments occurs within a relatively short period of time. Smooth muscle normally functions over a large length range; rapid onset of adaptation (plasticity) is therefore required. Our previous finding suggests that the recovery of isometric tension from length oscillation (which has disrupted filament organization) is complete within 30 min (12). The rate of recovery, however, depends on the frequency of activation; that is, the more frequently the muscle is stimulated, the faster the recovery. A similar observation was described for skeletal muscle stimulated periodically at short lengths (6). Adaptation of skeletal muscle to chronic shortening occurs by deleting sarcomeres in series. The process of adaptation is initiated by both a change in muscle length and repeated electrical stimulations. Adaptation occurs slowly (if at all) in unstimulated preparations set at short lengths in skeletal muscle (6). In rat soleus muscle, a period of 12 h of repeated stimulation is required for the adaptation process to complete (6). This is much longer than the 30 min required for airway smooth muscle stimulated at 5-min intervals (12).

An unanswered question is whether length adaptation occurs in the absence of stimulation, and, if it occurs, what is the minimal period of time required for the adaptation process to complete. The present study addresses these questions.

In patients who have chronic airway diseases such as asthma and chronic obstructive pulmonary diseases (COPD), prolonged active and passive shortening of smooth muscle contributes to airway narrowing. If adaptation of the muscle's L-T relationship to shorter length occurs under these circumstances, the ability of the smooth muscle to narrow the airways could be enhanced. The present study therefore addresses not only mechanisms of normal airway smooth muscle function but also mechanisms that may underlie pathological conditions of asthmatic or COPD airways.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Tissue preparation. A total of 12 rabbits (New Zealand White, 2-4 mo) were used for this study, 3 for each of the subprojects: 1) comparison between passively shortened and control, 2) comparison between passively lengthened and control, 3) tension recovery of control preparations after length oscillation, and 4) tension recovery of passively shortened preparations. Tracheas were removed from rabbits immediately after the animals were euthanized with CO₂ asphyxiation. The trachea was immersed in physiological saline 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. The trachea was then dissected free of connective tissue. Four specimens (each containing 2 cartilage rings, ~2.5-3 mm wide) were obtained from the central region of the trachea. The entire length (in situ) of the tracheal smooth muscle layer between the tips of cartilage in each double ring was measured while the cartilage rings were still intact. The cartilage ring was then cut open, leaving the smooth muscle preparation attached to cartilage at both ends. The epithelial layer and most of the connective tissue were removed from the smooth muscle layer. Aluminum foil clips were placed on the cartilage ends, precisely along the insertion points of the smooth muscle so that the distance between the edges of the clips was the entire length of the smooth muscle preparation. The dissected specimens were placed on wax at the bottom of a plastic container and covered in physiological saline solution. Pins were inserted through the aluminum clips to set the length of the smooth muscle. From each trachea, two specimens were used as control samples, i.e., the length of the smooth muscle was set at a reference length in which applied force was just enough to keep the muscle straight; another two specimens were either passively shortened or passively lengthened. To passively shorten a specimen, one end of the specimen was fixed by a pin and the other end of the specimen was allowed to be free of constraint. To passively lengthen a specimen, pins were placed such that the smooth muscle length was stretched to the in situ length. The prepared specimens were kept at 4°C in the physiological saline for 24 h before being mounted vertically in a tissue bath with one end fixed to a stationary hook and the other to a servo-controlled lever of a length-force transducer. The tissue bath contained physiological saline solution at 37°C that was bubbled with a 95% O₂-5% CO₂ mixture. The initial lengths of the muscle preparations were chosen to be a length at which there was no buckle or visible slacking of the preparations and at which the passive force was zero. The muscle preparation was equilibrated for 1-1.5 h at the initial length. During this time, the muscle was activated by electrical field stimulation (EFS) at 5-min intervals using a 60-Hz AC stimulator with platinum electrodes.

L-T measurements. After reaching equilibration at their initial length, the muscle preparations were stretched by 0.25-mm increments at 5-min intervals. An isometric contraction was elicited after each stretch. There was a 1-min delay between the stretch and stimulation to allow the passive tension to stabilize before activation. Both passive and plateau active forces were measured during each contraction. The stretches continued until the active force passed a maximum and declined to approximately the value at the initial length.

Tension recovery. To examine the rate and extent of tension recovery of passively shortened preparations, two groups of muscle preparations were set passively for 24 h at two lengths: the reference and the shortened length. First, an L-T curve for the preparation set at the reference length (first group) was obtained by stretching the muscle incrementally with 0.25-mm-length steps. A partial L-T curve for the shortened preparation (second group) was then obtained by stretching the muscle incrementally until it reached the L_o of the reference L-T curve (obtained in the first group). At this length, the preshortened muscle was on the descending limb of its L-T curve. While the muscle was maintained at this length, a series of isometric contractions was elicited at 5-min intervals to allow the muscle to adapt, and the active and passive tensions were recorded. The rate of tension recovery (or reversal of chronic adaptation) of the preshortened muscle was compared with the rate of tension recovery of a muscle subjected to a brief period of length oscillation at reference length (acute adaptation). The oscillation protocol was adopted from our previous study (12) to induce a temporary decrease in the potential for generating active tension, so the time course of tension recovery could be examined.

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

Data analysis. Both active and passive tensions were analyzed in relation to the length of the muscle preparations. A Weibull four-parameter function (SigmaPlot 5.0) was found to best fit the active force data, and the peak value was obtained from the fitted curve and taken as the maximal active force (P_o). L_o was determined as the length at which P_o occurred. Because the control and passively shortened preparations were from the same animals, paired t-test was used to compare the calculated L_o. Paired t-test was also used to compare L_o between control and passively lengthened preparations. One-way ANOVA was used to compare the L_o of the two control groups as well as the two passively altered groups. The time course of active force recovery was fitted to a two-parameter exponential growth equation, y = a(1  e^bx), where a and b are constants. The passive L-T curves were fitted with a two-component exponential decay function, y = a(e^bx) + c(e^dx), where a, b, c, and d are constants. For all statistical tests employed, P > 0.05 was not considered statistically significant. In each case, n is the number of muscle preparations used.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Length-tension measurements and L_o. The Weibull curve fit for all active force data yielded r² values between 0.96 and 0.99. The P_o values measured from the control and test preparations from the same trachea were similar. The mean and SD of P_o for the control preparations was 19.8 ± 2.5 mN. The estimated mean cross-sectional area of each preparation was ~0.3-0.5 mm². It was found that the L_o of passively shortened preparations was shifted to lengths shorter than the reference length from the same animal, and the L_o of passively lengthened preparations shifted to longer lengths. An example is shown in Fig. 1. Figure 1A shows a pair of muscle preparations adjacent to each other from the same trachea. One preparation was used as the control (set at the reference length), and the other was passively shortened for 24 h. Figure 1B shows a pair of muscle preparations obtained adjacent to each other from another trachea. One preparation was the control, and the other was passively lengthened for 24 h. Figure 1, top, shows that P_o values of these preparations were similar, and the lengths at which the P_o occurred were clearly different (passively shortened: 1.00 mm, control: 1.58 mm; passively lengthened: 2.43 mm, control: 1.64 mm). The same shift was repeated in the passive force measurements, which are shown in Fig. 1, bottom. The passive force remained at a level close to zero and increased at lengths longer than L_o. The passive force curve of the shortened preparation shifted to the left of its control, and the lengthened preparation shifted to the right. The amount of shift in passive force was consistent with that found in the active force. The means and SD of the L_o measurements are shown in Fig. 2. On average, the L_o values for passively shortened preparations were 28% shorter than the control. The preparations that were stretched to the in situ length for 24 h had an average L_o that was ~48% longer than the control values. One-way ANOVA showed that the L_o values of the two control groups were not different (P = 0.64) and that the L_o of passively shortened and passively lengthened groups were significantly different (P = 8.1e⁶). Paired t-test showed that the L_o of passively shortened and lengthened preparations were significantly different from their own control obtained from the same trachea (P = 0.018 and P = 0.0008, respectively).

View larger version (17K): [in this window] [in a new window]   Fig. 1.   Length-tension curves of rabbit trachealis. A: comparison between a passively shortened muscle preparation () and a reference preparation () from 1 trachea. B: comparison between a passively lengthened preparation () and reference preparation () from another trachea. Top: active force; circles are measured force at various lengths, and lines are the Weibull 4-parameter function fitted to data. Bottom: passive force at various lengths. See text for definition of reference length.

View larger version (14K): [in this window] [in a new window]   Fig. 2.   Means and SE (n = 6 muscle preparations) of the measured optimal length (Lo). Passively shortened and control (PS) preparations were from the same trachea. Passively lengthened and control (PL) preparations were from the same trachea.

Tension recovery. Passively shortened preparations were stretched to L_o of their paired controls. Active force increased with repeated stimulation at 5-min intervals. An example is shown in Fig. 3. The active L-T curve of the passively shortened preparation shifted to the left of the control (Fig. 3A). When the preparation was allowed to adapt at L_o of its paired control, the active force increased with time (Fig. 3). The measured isometric force from each contraction was converted to a fraction of total recovery. The means (n = 6 for the first 7 points and n = 5 for the last 3 points) and SE of the calculated fraction of recovery as well as the exponential fit are shown in Fig. 4. The level of active force at the start of recovery was 16.1 ± 6.5 mN (mean and SD). It took ~45-50 min for the active force to reach a plateau. The fully recovered level was 22.7 ± 10.8 mN (mean and SD). A two-parameter exponential growth function, y = a(1  e^bx), was found to best fit the time course of the active force, where y was the percentage of total recovery and x was time. The r² values of the fit ranged between 0.997 and 0.999. The b constant indicates the rate of recovery and was found to be 0.11 ± 0.04 (s¹). At the plateau, the recovered active force (22.7 ± 10.8 mN) exceeded the P_o of the paired controls (19.6 ± 9.9 mN). Paired t-tests showed that this difference was significant (P = 0.029).

View larger version (8K): [in this window] [in a new window]   Fig. 3.   Active force obtained from a pair of muscle preparations from the same trachea. , Reference preparations; , passively shortened preparations. A: active force vs. length. The preshortened preparation was allowed to adapt at reference Lo, resulting in increased tension with each stimulation. B: time course of tension recovery of the preshortened preparation at reference Lo. Line shows exponential 2-parameter fit.

View larger version (13K): [in this window] [in a new window]   Fig. 4.   Means and SE of fractional active force recovery vs. time. Active force is expressed as fraction of total recovery during adaptation at reference Lo. Dashed line and , recovery of control preparations at reference Lo after length oscillation. Solid line and , recovery of passively shortened preparations at reference Lo.

bx

dx

View larger version (7K): [in this window] [in a new window]   Fig. 5.   Passive force obtained from the same pair of muscle preparations in Fig. 3. , Reference preparations; , passively shortened preparations. A: passive force vs. length. The preshortened preparation was allowed to adapt at reference Lo. B: time course of passive force decline at reference Lo.

View larger version (10K): [in this window] [in a new window]   Fig. 6.   Means and SE of fractional passive force decline with time. Passive force is expressed as fraction of total force decline during adaptation at reference Lo.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Adaptation to length change in unstimulated muscles. The present finding of a bidirectional shift in the L-T curve of unstimulated airway smooth muscle implies that the subcellular components governing the L-T relationship can be restructured according to the length of the passively shortened or lengthened muscle cells. The underlying mechanism for this plastic behavior is not known. However, it is likely that the shift in L-T curve stems from variation in the number of contractile units in series, similar to the in-series sarcomere number change observed in skeletal muscle (6). The physiological significance of the shift is obvious: to optimize the contractile filament overlap. This is evident in Fig. 1. The length at which maximal tension occurs is different for different muscle preparations set passively at different lengths (for 24 h); the maximal tension, however, is the same at all lengths. It should be pointed out that the pair of preparations of muscle used as control and test came from adjacent rings in the same trachea; their intrinsic lengths are therefore close to be identical. The shift in L_o observed (Fig. 2) is therefore due to adaptation of the muscle to length change and not due to differences in their initial lengths.

Adaptation and tension recovery. Whereas the shift in the L-T relationship occurs slowly in relaxed muscle at 4°C, repeated activation at 37°C greatly speeds up the process, as shown in Figs. 3 and 4; a new P_o (with the associated L_o) is achieved in <50 min for the preshortened preparation (chronically adapted) and <30 min for the control preparation after an acute length perturbation. Our previous study (12) indicates that the frequency of muscle activation is positively correlated to the rate of tension recovery after a length change. The active tension recovery that occurs when the muscle is allowed to adapt at a constant length (Fig. 3A) follows a monoexponential process (Fig. 3B). We have previously observed a similar process of tension recovery in swine trachealis after length perturbation (12). The rate of tension recovery from acute length perturbation (b = 0.27) is faster than the rate of recovery from chronic adaptation (b = 0.11). This is likely due to the possibility that chronic adaptation may have resulted in a more permanent structural alteration that renders reorganization of the subcellular components a slower process.

L-T curve shift and implications in asthma and COPD. Chronic inflammation and the associated mediator and cytokines release seen in asthmatic and COPD patients result in a chronically narrowed airway due to smooth muscle activation and shortening. In addition, passive shortening of smooth muscle may occur in these conditions because of loss of lung elastic recoil and airway wall congestion as well as airway wall edema that effectively uncouples the muscle from lung recoil. Our in vitro results suggest that smooth muscle cells in chronically constricted airways may adapt to an abnormally short length. A direct mechanical consequence of this adaptation is the shifting of the L-T relationship, which in turn results in a maintained maximal force production and passive tension. According to Laplace law, the level of wall tension required to maintain a constant transmural pressure is proportional to the radius of the airway. In a constricted airway, less wall tension is therefore needed to maintain the same pressure. With muscle adaptation, the same maximal force produced by muscle at normal length can be generated by shortened muscle. This could create a catastrophic situation in which chronically narrowed airways can further constrict due to the mismatch of maximal force and the caliber of the airways. The shifting of the passive L-T curve further exacerbates the situation. Even without smooth muscle contraction, the leftward shift of the passive L-T curve in the shortened muscle makes distension of the airways more difficult.

Relation to early studies. When skeletal muscle is chronically lengthened or shortened it undergoes remodeling so that the length at which P_o is generated shifts to longer or shorter lengths, respectively (2, 7, 13). Skeletal muscle has been shown to achieve this adaptation by adjusting the number of sarcomeres in series in response to chronic static length changes (1, 5, 6, 8, 10, 11, 13). Marked changes occur in the size and shape of hollow organs that contain smooth muscle in certain physiological conditions (the uterus during and after pregnancy) and in pathological states (the bladder during chronic outflow obstruction). Smooth muscles have likely evolved a mechanism similar to that for skeletal muscle to accommodate large changes in length. It seems likely that smooth muscle has a series arrangement of identical contractile units (9). We have found in the present study that the maximal force generated by smooth muscle is independent of muscle length, similar to that found for canine trachealis (9). This suggests that, although there is variation in the number of contractile units in series due to chronic shortening or lengthening, manifested as shifts in L_o, the individual filament pairs are kept at their optimal overlap. This is similar to the situation in skeletal muscle in which the number of sarcomeres is adjusted to preserve the L-T relationship of individual sarcomere but shifting the whole muscle L-T relationship.