INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Current understanding of airway lumen narrowing in asthma emphasizes that muscle length and airway caliber are set by a balance of static forces. Force generated by airway smooth muscle statically is taken to be in mechanical equilibrium with the passive reaction force developed by the elastic load against which that muscle has shortened. Since both forces depend upon muscle length, the airway is presumed to accommodate itself to the muscle length at which these opposing forces have come into a static balance (1). The static balance corresponds to the point at which the static force-length characteristic of the activated muscle intersects the passive force-length characteristic of the load against which the muscle is contracting.

This mechanical equilibrium of static forces is sustained by the cyclic interactions of myosin with actin within the airway smooth muscle. In isometric muscle, with onset of a maximal contractile stimulus, myosin-actin cycling begins and the number of interactions (bridges) increases and approaches a plateau. During this process, rapidly cycling cross bridges convert progressively to slowly cycling latch bridges. This regulatory process eventually progresses to a steady state, and bridge dynamics are then said to have attained the latch state (4). In the latch state, the rate of bridge cycling has decreased to its smallest value attainable in maximally activated muscle (excepting the rigor state), the active force has increased to its maximum attainable value, and that value of the force defines one point on the static force-length characteristic of the muscle. Likewise, every point on the static force-length characteristic corresponds to an isometric steady state and the latch state.

Taken together, the concepts of a balance of static forces at the mechanical level and the underlying latch state at the molecular level have formed the foundation of our understanding of airway lumen narrowing (1, 2, 7, 8). These ideas are useful because they identify the static equilibrium length toward which activated airway smooth muscle would tend if given enough time.

But are the concepts of a static mechanical equilibrium and the latch state applicable in the setting of tidal loading, as occurs during breathing? The effects of tidal stretch on airway smooth muscle were first addressed by Sasaki and Hoppin (9) and later by Gunst and colleagues (10). These investigators demonstrated that imposition of tidal changes in muscle length inhibits force development. In this report, we extend this line of investigation to include stretch-induced changes in the muscle stiffness and, in particular, muscle hysteresivity,  (13). We have recently established that, like shortening velocity and the rate of actomyosin ATP utilization, the hysteresivity of airway smooth muscle is also governed by the rate of bridge turnover (14). Because it is easy to measure continuously throughout the contractile event, the hysteresivity provides a useful window on the rate of bridge turnover and its changes. Using this approach, we have studied isolated, maximally contracted bovine tracheal smooth muscle and assessed the way in which the interactions of myosin with actin are altered as a result of imposed tidal stretches.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In all to follow, we limit attention to airway smooth muscle that is maximally activated and operating in the vicinity of its optimal length, Lo.

Tissue Preparation

Bovine tracheas were obtained from a local slaughterhouse, and a section of four to five rings in the caudal to central region of each was stored in cold, phosphate-buffered saline for up to 24 h before use. The inner layer of connective tissue, adjoining cartilage, and outer connective tissue were carefully removed and muscle strips measuring 2 × 3 × 20 mm were dissected out. Each end of the tissue strip was glued (cyanoacrylate) to small brass clips attached to straightened steel music wires (0.37 mm in diameter). The muscle was suspended in a vertical glass tissue bath described previously (15). The upper wire was attached to a force transducer and the lower wire to a servo-controlled lever arm described below. The lower wire passed through a small hole at the bottom of the bath. A bead of mercury sealed the annular gap between the wire and the glass. The tissue bath was circulated with a Krebs solution (in mM: 118 NaCl, 4.59 KCl, 1.0 KH₂PO₄, 0.045 MgSO₄, 0.18 CaCl₂, 11.1 glucose, 23.8 NHCO₃; pH at 7.5 to 7.7) aerated with 95% O₂, 5% CO₂ and maintained at 37° C by a surrounding water jacket.

Coarse length changes were actuated by a micropositioner (Model 53-0311; Ealing). Small sinusoidal length excursions were imposed by the servo-controlled lever arm (Model 305B; Cambridge Technology). The length input signal was fed from a computer through a digital-to-analog converter and smoothed by a low-pass filter (10 Hz cutoff; Model 4113; Ithaco). The force was measured by a force transducer (Model FT.03; Grass) having an operating range of 1.0 kg, a resolution of 50 mg, and a compliance of 2.2 µm/g. Both the length and the force output signals were digitized at 66 Hz.

After equilibration for 1 h, the strip was brought to optimal length, Lo, in the standard manner using electric field stimulations adjusted for optimal response, beginning in the neighborhood of 50 V at 40 Hz, 1.5-ms pulse duration, for 30 s.

Experimental Protocol

Each bovine tracheal smooth muscle strip was stretched sinusoidally at a tidal strain amplitude , where  is L/Lo expressed as a percentage and L denotes the amplitude of length variations about Lo. Baseline data were collected with  at 0.25% and frequency f of 0.33 Hz for a duration of 100 s. At t = 100 s, acetylcholine (ACh; 10⁴ M) was introduced into the tissue bath, and  was kept unchanged at 0.25% until t = 400 s, at which time  was either maintained at 0.25% or increased to one of five stretch amplitudes: 0.5, 1, 2, 4, and 8%. At t = 900 s,  was decreased back to 0.25% for an additional 200 s. ACh was then washed from the bath for 10 min, and the protocol was then repeated at a different stretch amplitude.

For this range of tidal stretch, the resulting peak velocities during the stretch cycle ranged from 0.005 to 0.16 Lo/s; in preparations of this type, the maximal unloaded velocity of shortening after a quick release is in the range of 0.2 to 0.3 Lo/s.

Mechanical Measurements

We consider muscle subjected to small sinusoidal length variations about its optimum length, L(t) = Lo + Lsin(2ft) at frequency f and amplitude L. The total force in the muscle, FT(t), is the sum of the passive force, Fp, the active force, F(t), the elastic force, E(L(t)  Lo), and the frictional force, R(dL/dt), where E is muscle elastance and R is muscle resistance. Thus, FT(t) = F(t) + E(L(t)  Lo) + R(dL/dt) + Fp. We assume that during force development the active force changes approximately linearly with time over the duration of a single tidal stretch; when this trend is removed, a closed force-length loop results. The mean value of the total force over each tidal stretch was calculated. From the closed loops, the values of E, R, and  were computed on a loop-by-loop basis in the following manner. If D is the energy dissipated per period of imposed cyclic strain (i.e., area within the force-length loop) and F is the amplitude of the phasic force variation about F, we then use the following relations, which remain useful even when the loop becomes non-elliptical, which is indicative of nonlinear mechanical behavior (13, 15), E = (F/L)cos, R = (F/L)sin, and  = tan() where  = sin¹ (4D/FL). With sinusoidal length changes at radian frequency  (= 2f), then F() = (E + jR)L() = E(1 + j)L() where j =    1. The frictional (imaginary) part of the stress is proportional to R or, equivalently, E. Alternately,  may be regarded as the amplitude of the frictional force expressed as a fraction of the amplitude of the elastic force.

Inhibition of Stretch-activated Humoral and Neurogenic Pathways

Arachidonic acid metabolites are released in response to stretch and other forms of mechanical perturbation of tissues (16). In airway smooth muscle, the predominant cyclooxgenase metabolites of arachidonic acid are PGE₂ and PGI₂, both of which are bronchodilators (17, 18). Tidal stretch of tracheal smooth muscle also has the capacity to cause activation of neurons in the preparation and, by local reflexes, result in the release of neurotransmitters which may relax the tissue. In order to assess the role of prostanoid release and of reflexes in the changes of F, E, and  induced by tidal stretch, we treated paired tracheal strips with vehicle or with a combination of indomethacin (10⁶ M) and tetrodotoxin (10⁶ M) for 30 min prior to initiating the stretch protocol. Indomethacin is an inhibitor of the cyclooxygenase enzyme that converts arachidonic acid to prostanoids, while tetrodotoxin blocks fast sodium channels and thereby inhibits the generation of action potentials.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Force-Length Loops in the Steady State

We imposed about the optimum muscle length, Lo, a sinusoidal length fluctuation, L, at 0.33 Hz, as might occur during tidal breathing. The amplitude of the imposed tidal stretch expressed as a percentage of muscle optimal length ( = L/Lo) was maintained at a very small level, 0.25%, until the muscle was fully activated and the force-versus-length loop had attained a steady-state configuration.  was then increased to either 0.5, 1, 2, 4, or 8%. With this change, a new steady-state force-versus-length loop was established during the first few length oscillations (Figure 1). When  was small these loops were elliptical in shape, not very hysteretic, the chord slope was relatively steep, and the mean value of the active force was high. However, as  was increased, the loops became banana-shaped, more hysteretic, the chord slope became less steep, and the mean value of the active force fell. The force at peak muscle extension approximated isometric force for all values of , but the force at any lesser length was less than isometric force, and the more so when  was bigger.

View larger version (17K): [in this window] [in a new window]   Figure 1.   Steady-state loops showing force versus length for six amplitudes of tidal stretch,  = 0.25, 0.5, 1, 2, 4, or 8% of Lo, in fully activated bovine tracheal smooth muscle set to mean length of Lo. As  was increased, the loops became banana-shaped, more hysteretic, the chord slope became less steep, and the mean value of the active force fell.

Changes with Time and Approach to the Steady State when Tidal Stretch Is Negligible

To quantify the height, the slope, and the fatness of these loops and their changes in time, we calculated on a loop-by-loop basis the mean force over the stretch cycle, F, the stiffness (elastance), E, and the hysteresivity, , developed by the muscle. When  was maintained at 0.25%, addition of ACh to the bath (10⁴ M ACh at t = 100 s) caused E and F to increase monotonically and at similar rates (Figure 2A and B, solid lines). Both attained plateau values within 300 s. These changes of stiffness and force are thought to reflect increasing numbers of attached bridges (14).

View larger version (29K): [in this window] [in a new window]   Figure 2.   Time courses of force (A), stiffness (B), and hysteresivity (C ) in a representative tracheal smooth muscle strip. Solid trace in each panel corresponds to  maintained throughout at 0.25%. Broken lines correspond to runs in the same muscle subjected to graded increments of  from 0.25 to either 0.5, 1, 2, 4, or 8% over the time interval from 400 to 900 s. Control values are those measured at t = 400 s with  = 0.25%.

Changes of , by contrast, were prominently dissociated from those of E and F (Figure 2C, solid line). Addition of ACh caused  to follow a biphasic pattern in which it increased rapidly, peaked early in the contraction, and thereafter decreased slowly.  also attained a plateau value within about 300 s. Importantly, in a recent report, we established that like shortening velocity and actin-activated myosin ATP utilization,  is also a reasonable proxy for the rate of bridge turnover in activated muscle (14); as with baseline values of passive muscle force and elastance, the baseline value of muscle  is attributable to the passive mechanical properties of connective tissues. As such, these changes in  may be taken to reflect rapidly cycling cross bridges early in the contractile event converting to slowly cycling latch bridges later in the contractile event.

Effects of Increasing Tidal Stretch

was increased from 0.25% to a larger value at 400 s. Increased tidal stretch caused F and E to decrease promptly and  to increase promptly, within two to three stretches (Figure 2, broken lines). The greater the increase in , the greater were the resulting changes in F, E, and . Thereafter, the airway smooth muscle could be maintained in steady dynamically determined contractile states that differed demonstrably from those in isometric conditions. At t = 900 s,  was subsequently decreased back to 0.25%. This change caused F, E, and  to return slowly toward previously established values. Plateau values of F, E, and  were recorded at t = 900 s (i.e., 500 s after the increase of ) and replotted in a dose-response fashion (Figure 3), where dose was taken to be the amplitude of the tidal stretch and each was expressed as a percentage of its value measured at t = 400 s with  = 0.25%; these reference values approximate those in isometric steady-state conditions. As  was increased from this level, F and E progressively decreased and  progressively increased (Figures 3 and 4).

View larger version (18K): [in this window] [in a new window]   Figure 3.   Plateau values (t = 900 s) of F and E during tidal stretch expressed as a percentage of control (values measured at t = 400 s with  = 0.25%); values of  are absolute. The provocative stretch amplitude required to cause F or E to fall by half, or  to double, was slightly greater than 2%.

View larger version (11K): [in this window] [in a new window]   Figure 4.   When E is plotted versus , the data from Figure 2 defined a sequence of dynamically determined contractile states in which increasing tidal stretch pushed the muscle state down (less stiff) and to the right (more hysteretic).

Assessment of Stretch-activated Neurogenic and Humoral Pathways

The effects of tidal stretch on activated muscle strips (10⁴ M ACh) that were treated with tetrodotoxin and indomethacin were not different from the effects observed in strips that were treated with vehicle. There was a tendency for  of the strips treated with tetrodotoxin and indomethacin to be greater, but this tendency was not significant.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The principal finding of this report is that maximally contracted airway smooth muscle could be maintained in steady, dynamically determined contractile states for as long as the tidal stretches were sustained. These states were characterized by graded stretch-effect relationships in which the greater the tidal stretch amplitude, the greater were the departures of muscle force, muscle stiffness, and muscle hysteresivity from their values measured in conditions approximating the isometric steady state. The provocative stretch amplitude required to cause active force or muscle stiffness to fall by half, or hysteresivity to double, was slightly greater than 2%. Because muscle stiffness is a rough reflection of the numbers of actin-myosin interactions, and muscle hysteresivity is a rough reflection of the rate of turnover of those interactions (14), these results suggest that tidal stretch decreases the numbers of actin-myosin interactions and increases their turnover rate.

Gunst and colleagues (11, 12) and Rack and Westbury (19) showed that tidal stretches inhibit muscle force in smooth and striated muscle, respectively. There were several differences between the protocol used here and that of Gunst and colleagues that make detailed comparisons difficult, but as in these previous studies the data reported here indicate that the force at peak muscle extension approximates isometric force and that the force at any lesser length is smaller than the isometric force. With regard to hysteresis, we note that the activated state of airway smooth muscle has generally been considered to be a high-hysteresis state based on the studies of Sasaki and Hoppin (9), whereas we have shown here and previously (14) that maximally activated airway smooth muscle subjected to very small tidal stretches ( 0.5% of Lo) corresponds to a low-hysteresis state (Figure 2). This difference is reconciled by the fact that the studies of Sasaki and Hoppin correspond to rather large tidal stretches, which are now seen to be associated with larger values of  (Figures 1-3).

Mechanism

The dependencies of F, E, and  upon tidal stretch amplitude were uninfluenced by treatment of the tissues with tetrodotoxin and indomethacin. This observation rules out the importance of mechanisms based upon stretch-activated neural pathways or prostanoid release.

Another mechanism warranting consideration is stretch- induced remodeling of the cytoskeleton. Pratusevich and associates (20) and Gunst and colleagues (21) have demonstrated mechanical behavior that is suggestive of the idea that the focal adhesion complex and/or the cytoskeletal network in airway smooth muscle can remodel plastically during the course of force development. We believe that remodeling of this type is important and cannot be ruled out as a mechanism contributing to the responses reported here. However, we do not think that stretch-induced plastic remodeling could be the primary mechanism accounting for the observations reported here because of the rapidity, the magnitude, and the nature of the observed responses. First, when tidal stretch amplitude was suddenly increased, the effects on force, stiffness, and hysteresivity were essentially completed within two to three tidal stretches, or less than 10 s (Figure 2), whereas the remodeling events reported previously by Pratusevich and associates revealed themselves only over longer durations. Second, the onset of tidal stretches led to decrements of force and stiffness that were not only much faster but also much larger than those associated with remodeling events that have been reported previously (Figure 2). Finally, even if plastic remodeling of the cytoskeleton were able to account for stretch-induced changes in force and stiffness, both the temporal evolution of the hysteresivity and the stretch-induced change in the hysteresivity would remain unaccounted for and would then require their own ad hoc explanations.

This brings us to mechanisms at the level of the cross bridge, which could come into play in two ways. The first consideration is how muscle length modulates the maximum number of actin-myosin interactions that can be attained in the isometric steady state; changes in the number of interactions with changes of muscle length determine the shape of the static force-length characteristic and set the optimal length, Lo (22). The range of length excursions reported here represented deviations from Lo of 8% or less. The static force-length characteristic of smooth muscle is nearly flat over this range. Therefore, the shapes of the dynamic force-length loops (Figure 1) and the large force decrements (Figure 2) that occurred with the onset of tidal stretches cannot be attributed to traversing those parts of the static force-length characteristic where the static force is substantially smaller than it is at Lo. Similarly, the banana shape of these loops cannot be attributed to the well-known asymmetry of the Hill force-velocity curve on either side of the zero-velocity (isometric) point because, if it were, then the instantaneous force at both extremes of muscle length (where velocity is zero) would approximate the isometric force at that length, which was clearly not the case (Figure 1), and the extremes of force in the cycle would occur at the instants when the velocity of lengthening and shortening were greatest (both of which occurred at Lo), which was also clearly not the case.

The second consideration is the direct effect of tidal stretch on bridge dynamics. It is generally held to be true for all kinds of muscle that in the isometric steady state the rate of bridge cycling is constant and that if that steady state is disturbed sufficiently, then cross bridge dynamics will be altered. In striated muscle, it is well established that the inhibition of force and stiffness that occurs with the onset of sinusoidal stretches is attributable to the direct effect of tidal stretch upon bridge dynamics (19, 23, 24). In smooth muscle, the case is less clear (9). There continues to be substantial controversy surrounding Murphy's latch hypothesis governing cycling rate regulation even though the principal arguments against it, which were based on muscle energetics, seem now to have been reconciled (25, 26). According to the latch hypothesis, the population of myosin molecules is distributed among two attached species (the slowly cycling latch bridge and the rapidly cycling cross bridge) and two unattached myosin species. Attainment of an isometric steady contractile state implies that the distribution of myosin among its four species has come to a steady state. This isometric steady state is set by a dynamic equilibrium of seven rate processes, and when this equilibrium is attained the muscle is then said to be in the latch state.

Hai and Murphy (27) have shown that with judicious selection of the apparent rate constants for bridge attachment and detachment, the latch regulatory mechanism can be interpreted usefully within the context of Huxley's two-state sliding filament model (22). An exact mathematical solution of this model for the case of muscle subjected to sinusoidal length changes has recently been found by Jafari (28). Using the strain-dependencies of rate constants determined originally by Zahalak (24) in striated muscle and used more recently by Yu and coworkers (29) in smooth muscle, this solution demonstrates that as the amplitude of the tidal stretch is increased, the force and stiffness decrease and the hysteresivity increases, in a manner much as depicted in Figures 2 and 3, with stretch effects of the correct magnitude occurring over roughly the correct range of tidal stretch amplitude. Thus, a system as simple as Huxley's two-state sliding filament model expresses the essential stretch-dependent changes of F, E, and  that are reported here. On this basis, it seems reasonable to tentatively conclude that the dynamically determined contractile states depicted in Figure 3 are caused for the most part by a direct effect of tidal stretch upon bridge dynamics.

Jafari's analysis of Huxley's sliding filament model shows that these stretch-induced mechanical changes are attributable to disruption of the spatial distribution of myosin binding along the actin filament, and the departure of that distribution from the distribution that pertains in the isometric steady state. The distribution of myosin binding with actin is at every instant tending toward the distribution that would prevail in the isometric steady state, but imposed changes of muscle length cause an excess in the rate of detachment compared with the isometric steady state. Stretch-induced detachment events can come so fast compared with the rate of attachment that the static equilibrium distribution is never attained. Compared with the equilibrium distribution that prevails in the isometric steady state, with mechanical agitation the myosin binding distribution is pushed into disequilibrium and stays in disequilibrium for as long as tidal stretches are maintained. As a result force is inhibited, stiffness (i.e., bridge numbers) is inhibited, and hysteresivity (i.e., bridge turnover rate) is augmented compared with the isometric steady state. Being only a two-state analysis, however, Jafari's solution cannot address the questions of the partitioning of attached myosins between phosphorylated (rapidly cycling) and dephosphorylated (slowly cycling) species, the stretch-induced changes of that partitioning, or the degree to which these changes may account for the measured changes in muscle hysteresivity.

Functional Implications of Myosin Disequilibrium and the Paradox of Airway Lumen Narrowing

If for the moment we were to assume that muscle stretch scales isotropically as the cube root of lung volume change (30), then normal tidal lung inflations from FRC would correspond roughly to  of 4%, a sigh from FRC would correspond roughly to  of 12%, and inflation from FRC to total lung capacity would correspond roughly to  of 25%. It is interesting, therefore, that the provocative stretch amplitude required to cause active force or muscle stiffness to fall by half, or hysteresivity to double, is slightly greater than 2% (Figure 3). Thus, small tidal stretches, even smaller than would be expected with normal tidal breathing, are seen to be a potent endogenous relaxing mechanism, and one that expresses its greatest sensitivity to muscle stretch remarkably close to the range that would be expected to occur during spontaneous breathing. Accordingly, it is conceivable, if not likely, that normal individuals may live on the back porch of the stretch-effect relationship depicted in Figure 3, where force is profoundly inhibited (  4%), when their airway smooth muscle is maximally stimulated.

If tidal stretches even smaller than those expected with normal tidal breathing are so potent that they can inhibit most of the force developed in maximally contracted muscle, then one comes to the paradoxical conclusion that no appreciable airway narrowing ought ever to occur as a result of smooth muscle activation. In keeping with this notion, the diameter of the trachea and central bronchi does decrease substantially in both normal and asthmatic subjects during breathhold after cessation of tidal breathing (31). In addition, deep inspirations are known to reverse airway obstruction that is induced in healthy or asthmatic subjects (32). However, contrary to this notion, a deep inspiration fails to reverse obstruction that occurs spontaneously in asthmatics and most often exacerbates the degree of obstruction in these individuals (33, 34, 37). Regarding such observations, Fish and colleagues (34) noted that in spontaneous asthmatic obstruction there seems to be an intrinsic impairment of the ability of lung inflation to stretch airway smooth muscle. The asthmatic with spontaneous obstruction has lost a crucial defense mechanism that is otherwise potent. The salutary inhibition of bronchoconstriction by lung inflation is somehow disinhibited in that circumstance (34, 36).

Why do normal tidal breaths and even deep inspirations fail to inhibit airway narrowing in individuals with spontaneous asthmatic obstruction? No mechanism has been identified to explain the intrinsic impairment noted by Fish and colleagues. Several investigators have suggested that such behavior may be consistent with airway closure; once closed, the airway is difficult to reopen (3, 40, 41). Another possibility is that the stretch-related process that inhibits force development in normal muscle may be attenuated in the muscle of the asthmatic by some mechanism that is yet to be identified. Yet another possibility is the presence of a myogenic response in asthmatic airway smooth muscle, although there is little evidence for that possibility (42). The last and simplest possibility, however, is that the smooth muscle of the asthmatic finds itself operating at amplitudes of tidal stretch that are systematically smaller than does smooth muscle of the normal individual (Figure 3).

A plausible case can be constructed to support this last possibility. In recent years, inflammatory remodeling of the peribronchial adventitia has been demonstrated in the asthmatic lung, and many investigators have argued that adventitial thickening acts to decrease the static parenchymal load against which airway smooth muscle must shorten (2, 7, 43, 44). If so then, similarly, adventitial thickening would tend to diminish the tidal forces to which the smooth muscle is subjected. The tidal stretches of airway smooth muscle would be less, and the effectiveness of tidal lung inflations in inhibiting force development would be diminished accordingly, displaced leftward in Figure 3. With progressive adventitial thickening, therefore, the load against which the muscle shortens becomes diminished (2, 7, 44) while the force generated by the muscle becomes disinhibited (Figure 3). So, the progressive adventitial thickening that occurs in asthma has two edges, like scissors. Caught between rising muscle force and falling muscle load, the airway lumen can become severely compromised.

If by adventitial thickening, or by other mechanisms, maximally activated muscle in the asthmatic lung should find itself subjected to systematically smaller tidal forces than does muscle in the normal lung, then the effects of these differences would be expected to be reinforced, if not amplified, by the dynamic properties of the muscle itself (Figure 3). We reason as follows. Should the maximally contracted muscle find itself operating on the stiff part of its E- relation (say  < 2%; Figure 3), then with each tidal breath the muscle would substantially impede its own tidal stretch because it is so stiff (45, 46). In this case, the tidal stretch would be small, the active force would be large, and the underlying contractile state would approximate the one that pertains in steady isometric conditions. Thus, it may be useful to think of the muscle in that circumstance as being stuck, or frozen as it were, because its stiffness is so great. However, if maximally activated muscle should somehow find itself operating on the more compliant part of its E- relation (say   4%; Figure 3), then with each tidal breath the muscle could muster little impedance to its own stretch because its stiffness is so small. Smaller stiffness logically factors larger tidal stretches, and larger tidal stretch causes smaller stiffness (Figure 3). Or as Malmberg and associates (45) said, "a constricted muscle that is partially broken up [by lung inflation], may be more susceptible to further break up." Therefore, tidal stretch could act as its own catalyst, and such a state would tend to be self-reinforcing. As a result the tidal stretch would be large, the muscle force and stiffness would be inhibited substantially, and the underlying contractile state would be in disequilibrium (in the sense described above), dynamically determined, and far from the one that pertains in steady isometric conditions (Figure 3). Muscle length would then be determined by its dynamic properties, rather than its static properties, and the muscle state would not be found on the static force-length characteristic. We put forward the hypothesis that spontaneous obstruction in asthma may correspond to the former state of affairs ( < 2%) and that induced obstruction in healthy individuals may correspond to the latter ( > 4%). If true, this might explain the ability of lung inflation to stretch airway smooth muscle in some circumstances and the intrinsic impairment of that ability in other circumstances (32). A corollary of this hypothess is that the associated changes of hysteresivitysmall when stretch is small and larger when stretch is largerare of the right sign to account for the effects of volume history noted by Lim and colleagues, who attributed the effects to changes of parenchymal hysteresis rather than airway smooth muscle hysteresis (33, 37).

In the broader context, however, it is important to emphasize that these concepts cannot explain the asthmatic reaction itself and are intended only to highlight some of the interesting downstream events, at the level of bridge dynamics, that might condition the effects of that reaction. As described below, these events are expected to depend upon the rate of bridge cycling in relation to the pattern of tidal lung inflation.

Speculations on the Relation between Myosin Disequilibrium and Airway Hyperreactivity

The maximal shortening velocity, Vmax, is found to be increased in allergen-sensitized airway smooth muscle of dogs, mice, and humans and is greater in the developing lung than in the mature lung, although the role of maturation may differ between species (47). Importantly, airway hyperreactivity is associated with each of these circumstances, but no differences in isometric force generation capacity of the muscle are found. Therefore, it is impossible to explain airway hyperreactivity on the basis of muscle dysfunction if muscle length is set by a mechanical balance of static forces, where muscle shortening has been completed, muscle velocity is zero, and steady isometric conditions pertain. For this reason, the search for the cause of airway hyperreactivity has focused on extra-muscular factors (1, 52). However, in the context of the disequilibrium of airway smooth muscle, Vmax emerges naturally as a central determinant of muscle length because it is an index of the intrinsic rate of cross bridge cycling (53). Increased Vmax suggests an increase in the rate of cross bridge cycling. If, in the healthy challenged subject, tidal changes in muscle length can come so fast that static equilibrium conditions are never attained, then, from the point of view of airway muscle with increased intrinsic rates of cross bridge cycling (as in allergen-sensitized or some immature muscles), tidal changes of muscle length would not seem to be coming quite so fast. Therefore, myosin binding kinetics could approximate those in static conditions that much more closely, implying greater capacity for force and stiffness development in the dynamic circumstances that pertain during normal tidal breathing. So, perhaps airway smooth muscle dysfunction does contribute to airway hyperresponsiveness after all (52).

Finally, excessive narrowing of the airway lumen that is responsive to bronchodilators has been reported to be associated with cervical spinal cord injury, obesity, normative aging, emphysema, cystic fibrosis, scoliosis, and sleep (nocturnal asthma) (54). Further, the excessive narrowing of the airway lumen that can follow hyperpnea, especially breathing cold dry air, is entirely suppressed during the hyperpnea itself. Finally, it has been postulated that small airway occlusion may play a key role in sudden infant death syndrome (59). All of these circumstances have in common either an altered pattern of tidal lung inflations and/or reason to suspect diminished effectiveness of those inflations in stretching airway smooth muscle. This raises the question of the degree to which the mechanisms discussed here may come into play in those circumstances and explain seemingly distinct hyperreactivity phenotypes.

Conclusions

The data presented here suggest that tidal lung inflations have the potential to modulate profoundly the behavior of maximally contracted airway smooth muscle. We suggest that escaping steady-state isometric conditions, versus failing to do so, may be a major mechanism distinguishing the behavior of activated airway smooth muscle in the normal lung from that in the spontaneously obstructed asthmatic lung. These differences are a manifestation of the dynamic load against which the muscle is operating, as much as that of the muscle itself. Therefore, under particular circumstances, airway smooth muscle that is perfectly normal could become stiff enough that it becomes virtually frozen in isometric steady-state conditions, thereby explaining the observations of Fish and colleagues (34), but in other circumstances it would not. Abnormality of the muscle, if any, might have important consequences, but abnormality of the muscle would not be a prerequisite for abnormal muscle behavior and excessive airway narrowing.

To us, the most impressive evidence in this regard is the recent study of Skloot and associates (60, 61), who demonstrated in normal subjects that a rather modest perturbation in the pattern of tidal lung inflation, namely, using normal lung volumes and normal tidal volumes but with prohibition of deep lung inflations, causes not only a precipitous change in responsiveness of these individuals to contractile challenge but also causes a marked attenuation of the ability of these normal individuals to reverse obstruction with subsequent deep inspirations, a behavior observed previously only in spontaneously obstructed asthmatics. In this connection, Warner and Gunst (62) showed that tidal inflations inhibit airway narrowing in challenged canines, but that tidal breathing below transpulmonary pressures of 7.5 cm H₂O was not adequate to prevent airway closure. Taken together, our findings, those of Skloot and associates, and those of Warner and Gunst lead us to suspect that tidal breathing alone may be almostbut not quiteadequate to disrupt bridge dynamics in vivo. However, when tidal lung inflations are leavened by occasional deep inspirations, which occur in humans at a rate of no less than 10 times per hour during spontaneous breathing (36, 63), bridge dynamics may then become sufficiently disrupted that a self-reinforcing dynamically determined contractile state does emerge, a state in which the myosin binding distribution is in disequilibrium and far from the one that pertains in steady isometric conditions. Therefore, in the studies of Skloot and associates, we suspect that prolonged prohibition of deep inspirations in the presence of contractile stimuli allowed the myosin binding distribution in these normal subjects to collapse toward the one that pertains in steady-state isometric conditions, which in turn allowed the muscle to become so stiff that it was refractory to the influence of subsequent deep inspirations.