|Home | About | Journals | Submit | Contact Us | Français|
Breathing is known to functionally antagonize bronchoconstriction caused by airway muscle contraction. During breathing, tidal lung inflation generates force fluctuations that are transmitted to the contracted airway muscle. In vitro, experimental application of force fluctuations to contracted airway smooth muscle strips causes them to relengthen. Such force fluctuation–induced relengthening (FFIR) likely represents the mechanism by which breathing antagonizes bronchoconstriction. Thus, understanding the mechanisms that regulate FFIR of contracted airway muscle could suggest novel therapeutic interventions to increase FFIR, and so to enhance the beneficial effects of breathing in suppressing bronchoconstriction. Here we propose that the connectivity between actin filaments in contracting airway myocytes is a key determinant of FFIR, and suggest that disrupting actin-myosin-actin connectivity by interfering with actin polymerization or with myosin polymerization merits further evaluation as a potential novel approach for preventing prolonged bronchoconstriction in asthma.
Excessive contraction of airway smooth muscle (ASM) leads to bronchoconstriction and airflow obstruction during an asthma attack. In nonasthmatic airways, the ASM is found in thin bundles arranged essentially circularly around the airway, likely providing structural support. In normal lungs, ASM also likely assists in regulation of gas exchange, mucus clearance, defense mechanisms, and coughing, though it is uncertain whether any of these roles is required for normal lung function (1). Indeed, it may be that airway smooth muscle is a vestigial remnant of lung development (2), during which its periodic contractions might raise luminal pressure and thereby help to advance the lung buds (3). In the airways of individuals with asthma, the smooth muscle layer becomes thicker by increasing cell size and/or cell number. Such ASM over-accumulation is thought to facilitate the excessive bronchoconstriction that occurs characteristically in asthma. Furthermore, evidence connects smooth muscle in the pathogenesis of airway inflammation and remodeling (4, 5).
Smooth muscle contracts when filaments (thick) of myosin motors hold and ratchet actin filaments (thin) in opposite directions within the cell cytoplasm (Figure 1). These contractile actin filaments are oriented roughly along the long axis of each myocyte, and are mechanically interconnected by co-anchorage within dense bodies, or by inter-filament linkage by thick filaments; some thin filaments are also anchored to dense plaques at the cell membrane and (through integrin complexes) to surrounding tissue matrix. Thus, myosin-mediated translation of thin filaments in opposite directions shortens individual myocytes along their long axis, and in concert they shorten the muscle bundles in which they share long axis orientation.
Because the airways are embedded within the lung parenchyma, the load against which muscle bundles contract is determined in part by lung inflation; thus, breathing imposes a fluctuating load against which shortening must occur or must be maintained. While the steady mean load against which airway smooth muscle contracts partially determines its extent of shortening during maximal contractile stimulation, the influence of superimposed load fluctuations (such as those imposed by breathing) appears to be of even greater importance. For example, Shen and coworkers demonstrated that repeated single boluses of intravenous methacholine caused bronchoconstriction that varied considerably according to the depth of tidal breathing (at fixed frequency) in mechanically ventilated rabbits (6). Airway resistance rose considerably in response to methacholine administration when tidal volume was zero, whereas the bronchoconstrictor response was progressively attenuated when the same methacholine dose was given during mechanical ventilation with higher and higher tidal volumes; indeed, bronchoconstriction was prevented essentially completely by mechanical ventilation with large (20 ml/kg) tidal volumes. Hyperpnea similarly suppresses bronchoconstriction in other animals and in humans (7–10), and accounts for the delayed onset of bronchoconstriction induced in subjects with asthma by exercise (which provokes hyperpnea) or by eucapnic voluntary hyperventilation of dry air, until after cessation of exercise or hyperpnea. Thus, even in asthma, deep breathing functionally antagonizes bronchoconstriction, though studies (11–15) suggest that the degree to which breathing antagonizes bronchoconstriction may be abnormally reduced in subjects with asthma.
Superimposition of force fluctuations (that mimic the load fluctuations created by breathing; frequency 12/min) upon a submaximal steady load against which ASM strips that had been isotonically contracted results in their partial relengthening in vitro (Figure 2) (16–20). Such “force fluctuation–induced relengthening” (FFIR) of shortened ASM strips can occur even during continued maximal contractile stimulation (e.g., continued high concentration acetylcholine exposure) and continued maximal contractile activation (reflected in high level phosphorylation of the regulatory 20-kD myosin light chain). Clearly, FFIR is a different phenomenon from muscle relaxation, even though both result in relengthening of the shortened muscle. Furthermore, FFIR of maximally contracting ASM strips is a physiologically regulated phenomenon, for the extent of relengthening, and the time course over which it occurs varies considerably with the mode of contractile activation (e.g., acetycholine versus KCl exposure; see below) and with manipulation of intracellular signaling pathways (18).
We propose that the connectivity between actin filaments constitutes a major determinant of the magnitude of FFIR exhibited by contracted ASM strips in vitro. As noted above, actin filaments may be physically connected within ASM cells by virtue of their simultaneous attachment to opposite faces of a common myosin filament (Figure 1). We and others have gathered evidence that conditions which modulate the strength or nature of such actin filament–myosin filament–actin filament connectivity dramatically influences the magnitude and time course of FFIR of contracted airway smooth muscle.
It seems likely that forcibly extending airway myocytes with short actin filaments could cause sufficient sliding of the filaments relative to each other as to lose a zone of “overlap” in which both filaments could be connected together through their common attachment to a single myosin filament. Were such loss of connectivity to happen as a result of stretching during a force fluctuation, then internal rearrangement and reconnection of the actin and myosin filaments might occur in such a way as to reduce the force-generating capacity of the contractile apparatus, as illustrated in Figure 3A (21). Essentially, the rearrangement of actin and myosin filaments in a forcibly lengthened muscle cell with short actin filaments might require that previously parallel force-generating myosin motors now assume a series distribution, which reduces their concerted force-generating capacity, thus reducing the myocyte's ability to resist lengthening in response to force fluctuations and increasing FFIR. In contrast, myocytes with long actin filaments might not undergo parallel-to-series rearrangement of their myosin filaments, and so might not lose force-generating capacity with which to restore myocyte length after stretching (Figure 3B).
In support of this notion are results from two studies in which different pharmacologic interventions that might have shortened actin filaments also increased FFIR. Lakser and colleagues (18) showed that inhibition of p38 mitogen-activated protein kinase (MAPK) with SB203580 increased FFIR exhibited by maximally activated bovine tracheal smooth muscle strips shortened against 32% (mean) of their maximal force (Figure 2). p38 MAPK proteins are expressed in ASM and become activated during acetylcholine (ACh)-induced contraction (and in response to diverse stress stimuli, including hypersomolarity, cooling/rewarming, and IL-1β or TNF-α exposure). Acting through its downstream targets and effectors, MAPKAP kinase 2/3, activated p38 MAPK stimulates phosphorylation of HSP27 on three serine residues (22). Nonphosphorylated HSP27 is thought to bind to and cap the barbed ends of actin filaments, thereby preventing the addition of monomers and lengthening of actin filaments (23). Phosphorylated HSP27 can no longer bind to actin filaments, and their barbed ends are thus available for actin monomer addition and consequent filament lengthening (24). In the study of Lakser and coworkers, pharmacologic p38 MAPK inhibition likely also blunted HSP27 phosphorylation, thereby promoting actin filament capping and shortening actin filaments. More direct evidence that actin polymerization influences FFIR was gathered when Dowell and colleagues (25) demonstrated that sequestration of actin monomers with latrunculin B augmented FFIR in contracted canine tracheal muscle strips. Actin filament length is ordinarily determined by the balance of addition and loss of actin monomers from actin filaments. By sequestering actin monomers, latrunculin treatment should have prevented the addition of G-actin to actin filaments, thereby shortening these filaments (though actin filament length was not directly measured in that study). Thus, two different interventions that both should have resulted in shorter actin filaments, and so should have changed the susceptibility of actin filament–actin filament connectivity to force fluctuations, each enhanced the magnitude of FFIR.
It is the side-polar myosin filaments that actually connect actin filaments to one another in contracting ASM, through myosin-actin crossbridge interactions along both sides of the myosin filament. The number of crossbridge attachments in the cross-section of the tissue determines the force supported by a muscle. For a given contractile activation state (i.e., a given likelihood that each myosin head is attached to actin), a longer myosin filament with more myosin heads should more tightly interconnect its two flanking actin filaments, by virtue of a greater number of crossbridge attachments to each actin filament. Thus, longer myosin filaments should produce stronger actin filament–actin filament connectivity than would shorter myosin filaments. Importantly, myosin filament length is not fixed (26), but rather is dynamically determined, changing even during the course of a single contraction (27, 28). In addition, myosin filament length appears to be influenced by recent muscle length history (e.g., length oscillation of relaxed ASM shortens myosin filaments) (29), by the length of actin filaments which seem to provide a template for myosin polymerization (30–32), and by the phosphorylation state of the 20-kD regulatory myosin light chain (MLC20) (33–38). We exploited the latter to obtain indirect evidence suggesting that myosin filament length modulates FFIR. We chose to compare FFIR on ACh- and KCl-elicited contractions because we wanted to be able to assess differences in contraction, FFIR, and selected protein activation in canine tracheal smooth muscle (TSM) strips that had been activated by two diverse mechanisms: G protein–coupled receptors and voltage-dependent calcium channels, respectively. While either ACh or KCl can stimulate substantial force generation and ASM shortening, each agent does so employing signaling pathways that only partially overlap with those activated by the other. Consequent to these different signals, greater MLC20 phosphorylation is seen in maximally ACh-stimulated than in maximally KCl-stimulated tissues (Figure 4B). We did not use lesser concentrations of ACh to elicit after-loaded isotonic contractions in the TSM strips because of the difficulty in “titrating” equivalent levels of isometric, isotonic, or after-loaded isotonic activation among several strips from the population of canine tracheas. Lesser concentrations of ACh would be on the linear portion of the sigmoidal concentration–response relationship and subject to greater variability. Also, G protein–coupled receptor activation with ACh at a lower concentration would still activate MEK and p38 MAPK pathways that promote contractile filament assembly.
The time course of FFIR is dramatically different between ASM strips that have been contracted with ACh versus strips contracted with KCl. Figure 4A shows the length changes typically observed when canine trachealis strips stimulated with 10−4 M ACh or 43 mM KCl-substituted Krebs-Henseleit solution are shortened isotonically against 32% Fmax steady load (as determined during isometric contraction activated by each agonist, respectively), then subjected to superimposed load fluctuations (±16% Fmax) during continued contractile activation. Each muscle strip shortens by about half of its reference length against the steady load. Application of load fluctuations to the ACh-contracted muscle strip induces a small immediate increase in length, followed by progressive relengthening toward an asymptote thereafter. In marked contrast, the KCl-contracted strip exhibits a dramatic immediate increase in length upon application of load fluctuations, and settles at a final length soon after the abrupt relengthening induced by force fluctuations.
A mechanical system that could mimic the immediate relengthening response to force fluctuations in KCl-contracted ASM strips is depicted in Figure 5. In this schema, the cell at the top of Figure 5 presents intramyocyte conditions after isotonic shortening and just before superimposition of force fluctuations. A pair of nodes mechanically anchoring the contractile apparatus (akin to dense bodies or dense plaques) have been pulled toward each other during contraction, and end up connected by two types of mechanical pathways. One is shorter and taut, and actually supports the load against which contraction has occurred; the other is longer and slack, and at this point in time is not actually supporting much load. Imagine further that the tensile strength of the shorter connector is relatively low (think “wet spaghetti”), so that the load it now bears is near its breaking point. In contrast, the tensile strength of the longer connector is great (think “steel chain”)—more even than the greatest load that it could ever experience within the myocyte. Now let us superimpose force fluctuations upon the steady load that the cell at the top of Figure 5 had shortened against. As the load fluctuates above the previously steady level, it exceeds the tensile strength of the wet spaghetti path (which had borne the entire load by virtue of its shorter length), and the wet spaghetti path breaks. No longer held in its shortened configuration by the wet spaghetti, the myocyte quickly relengthens until the steel chain path is fully extended, whereupon it supports the entire external load and prevents further relengthening. The integrated system thus snaps from a steady shorter length during isotonic contraction to a new steady but longer length during load fluctuation, in a pattern that mimics the relengthening of KCl-contracted canine trachealis muscle. Our model depicts contractile units within an isolated cell. It is conceivable that through adherens junctions (force transmission sites formed between dense plaques of adjacent cells), contractile units in different cells may similarly interact.
We speculate that because there is relatively low MLC20 phosphorylation in KCl-contracted ASM, the distribution of lengths of myosin filaments interconnecting actin filaments is skewed toward predominance of shorter myosin filaments (Figure 6, top), which as discussed above might result in weaker actin filament–myosin filament–actin filament connectivity. A contractile pathway containing such weak actin-myosin-actin connections would break when the tensile load imposed (e.g., during force fluctuations) exceeded the connection strength (39). Contractile paths with low tensile strength (due to their containing actin-myosin-actin connections through short myosin filaments, as shown in red in Figure 6) would behave like the wet spaghetti paths in Figure 5. Upon their rupture during force fluctuations, the muscle would lengthen quickly until pathways composed of only stronger actin-myosin-actin connections (i.e., through longer myosin filaments, as shown in blue in Figure 6) support the entire load and prevent further relengthening. We imagine that the very minor immediate relengthening that occurs in ACh-contracted ASM strips subjected to force fluctuations (Figure 4A) reflects a predominance of longer myosin filaments (whose polymerization had been promoted by increased MLC20 phosphorylation; Figure 4B), as shown in Figure 7. In such muscle, virtually all the paths are “kryptonite chains,” due to the very strong actin filament–actin filament interconnections by long myosin filaments. Perhaps the prolonged FFIR that occurs in ACh-contracted muscle strips reflects the sequential breakages of the weakest of the strong pathways, each fracture allowing for tiny length increments that accumulate slowly (39). Alternatively, perhaps other mechanisms, such as those influenced primarily by actin filament length, account for the slow relengthening phase.
If it occurs in vivo, which seems extremely likely, then force fluctuation–induced relengthening of contracted airway smooth muscle would have the effect of reversing bronchoconstriction. Thus, intervention to enhance FFIR could represent a novel therapeutic approach for airflow obstruction in asthma, one that does not rely on relaxing airway muscle (the principal goal of acute reliever medications) or subduing airway inflammation (the principal goal of long-term controller medications).
As discussed above, it seems likely that strong connectivity between actin filaments reduces force fluctuation–induced relengthening of contracted ASM. Thus, disrupting actin-myosin-actin connectivity seems worthy of exploration as a potential treatment for asthma. We suggest two approaches for disrupting actin-myosin-actin connectivity: (1) shorten actin filaments; and (2) shorten myosin filaments. Actin filament length is regulated physiologically through the balance of addition and loss of actin monomers and severing or stabilization of actin filaments, as we reviewed previously (21). These processes are not unique to airway smooth muscle. Given the ubiquitous importance of actin filaments in a wide range of cell types, it might be challenging to design an intervention that shortens actin filament length selectively within airway smooth muscle without disrupting other important actin functions in the lung, such as maintenance of airway epithelial or vascular endothelial barrier integrity. In contrast, smooth muscle myosin differs structurally from skeletal or cardiac myosin, and it may be easier to inhibit smooth muscle myosin polymerization selectively. Although we are not aware of drugs developed specifically for this purpose, monoclonal antibodies directed toward specific parts of the myosin II tail prevented or even reversed myosin polymerization in vitro (40, 41). Perhaps small-molecule drugs could be designed to achieve this goal.
Supported by NIH Grants HL007605, HL079398, and HL086604.
Conflict of Interest Statement: T.L.L. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. M.L.D. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. O.J.L. is an investigator on a research grant supported by AstraZeneca and administered by the University of Chicago. W.T.G. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. J.J.F. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. C.Y.S. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. R.W.M. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. J.S. has received or will receive consultant fees of $1,000 in 2005 from Critical Therapeutics, $1,000 in 2006 from Tanox, $1,600 in 2006 from Merck, $2,700 in 2007 from AstraZeneca, $3,500 in 2007 from Genentech, $6,000 in 2008 from Cytokinetics (with a similar amount expected in 2009), and $2,500 in 2008 from Sepracor. He has also served as principal investigator on a grant of $125,031 from AstraZeneca to the University of Chicago in 2008, with similar additional funding anticipated during 2009.