Emergence of airway smooth muscle functions related to structural malleability

Abstract

The function of a complex system such as a smooth muscle cell is the result of the active interaction among molecules and molecular aggregates. Emergent macroscopic manifestations of these molecular interactions, such as the length-force relationship and its associated length adaptation, are well documented, but the molecular constituents and organization that give rise to these emergent muscle behaviors remain largely unknown. In this minireview, we describe emergent properties of airway smooth muscle that seem to have originated from inherent fragility of the cellular structures, which has been increasingly recognized as a unique and important smooth muscle attribute. We also describe molecular interactions (based on direct and indirect evidence) that may confer malleability on fragile structural elements that in turn may allow the muscle to adapt to large and frequent changes in cell dimensions. Understanding how smooth muscle works may hinge on how well we can relate molecular events to its emergent macroscopic functions.

if life is a collective property of actively interacting molecules (43), then the same can be said about smooth muscle function. Smooth and striated muscles share many common properties; both tissues are highly specialized to generate force and perform associated mechanical functions. However, many of the emergent behaviors of smooth muscle differ dramatically from those of striated muscle. The most noticeable differences have to do with malleability of the contractile apparatus and supporting cytoskeletal scaffolding. From the need to adapt to large and frequent changes in cell dimensions, smooth muscle appears to have evolved a remarkable malleability that is readily modifiable by both mechanical and chemical stimuli. As a result of this malleability, the dynamic mechanical properties of the cell have no characteristic scales of length or time. Even the classical concept of optimal muscle length is now understood to be largely a myth and has been replaced by the concept of length adaptation.

In this minireview, the known emergent behaviors of smooth muscle relating to structural malleability are described, although the specific molecular interactions that give rise to these emergent behaviors remain largely unknown. However, some known properties of the elementary structures of smooth muscle, such as the contractile and cytoskeletal filaments, are thought to contribute to these emergent behaviors; the properties of these elementary structures are discussed in the second half of the review. Although this review is based on literature in the area of airway smooth muscle (ASM), the phenomena and behaviors described here are not unique to ASM and are common with many types of smooth muscle and nonmuscle cells.

RESPONSE OF ASM TO MECHANICAL PERTURBATION

Length adaptation.

The characteristic relationship between active force and sarcomere length in striated muscle can be traced back to the interaction of myosin cross bridges with actin filaments and the length dependence of that interaction within the sarcomere (14). Because the exact structure of the contractile unit in smooth muscle (akin to the sarcomere in striated muscle) is not known, the length-force relationship of smooth muscle cannot be linked to specific structures within the contractile unit. The length-force relationship of smooth muscle likely stems from interactions of myosin cross bridges with actin filaments (18), as in striated muscle. However, the emergent length-force properties of these two types of muscle are different, as shown in Fig. 1A, suggesting that the length-dependent cross-bridge interaction with actin filaments is different in smooth muscle. This difference can be dissected into at least two aspects; the first is that the shapes of the length-force curves (at least the ascending portion of the curves) are different, and the second is that in smooth muscle the length-force relationship cannot be represented by any single curve. Mechanisms underlying these differences are likely different and are discussed separately as follows.

Fig. 1.

A: length-force relationship of smooth muscle. Gray lines depict the same relationship in striated muscle, redrawn from Gordon et al. (14) as a reference, with permission from John Wiley and Sons, Ltd. Solid lines are linear fits of data from airway smooth muscle (ASM) adapted to different lengths. The red arrows with dotted lines depict sequences and directions of changes in force after shortening from the reference length and after stretch from the reference length during the process of length adaptation. See text for more detail. B: schematic of a smooth muscle contractile unit. Dotted arrows indicate the direction of movement of thin filaments (and their associated dense bodies) during active shortening.

The shape of the length-force curve in smooth muscle should reflect the structural constraint imposed by the contractile unit, similar to the situation in striated muscle. Based on the observed linear relationship of the individual length-force curves shown in Fig. 1A, a simple model can be constructed that predicts such a relationship (Fig. 1B) if two assumptions are made. 1) In a fully adapted smooth muscle (to any particular length), the myosin (thick) filament overlaps the actin (thin) filaments completely and spans the whole distance from dense body to dense body. 2) Once slid over the dense bodies the ends of the myosin filament no longer contribute to force generation. With these assumptions, the model predicts a linear length-force relationship because the overlap between the thin and thick filaments decreases linearly as the muscle length (or the length of the contractile unit) decreases.

As shown in Fig. 1A, the linear length-force relationship of smooth muscle can be shifted in a process called length adaptation. As indicated by the red dotted arrows (Fig. 1A), if the length of the muscle is suddenly shortened, for example, from an arbitrarily chosen reference length (L_ref) (at which the muscle has been fully adapted) by 25%, the force will decrease in proportion to the amount of length change; if the muscle is allowed to “recover” (usually over a period of 30 min during which the muscle is briefly stimulated at 5-min intervals) at the shortened length, the force returns to the maximal value before the length change (18, 38), and in the process shifts the whole length-force curve to the left. This is commonly referred to as length adaptation (1). Length adaptation works in both directions, that is, if the muscle is stretched (and full force recovery is allowed to occur), the length-force relationship will shift to the right (Fig. 1A).

Several lines of evidence suggest that length adaptation accompanies structural changes that alter the number of contractile units, especially the units in series spanning the cell length (38, 46, 25). Maximization of the overlap between the thick and thin filaments at different cell length is speculated to be the result of structural remodeling, as supported by the evidence of recovery of maximal force at different adapted lengths. The model also explains the increase in velocity and power output when the muscle is adapted at a longer length while preserving length independence of maximal active force (18, 25, 38).

Force-fluctuation induced relengthening.

The extent of ASM shortening in vivo is not determined by the length-force relationship described above, but by a complex interaction among several variables. Many of these variables come into play because the muscle resides in a mechanical environment that is dynamic: the muscle is constantly subjected to periodic stretches due to fluctuations in the transmural pressures in the airways because of the action of breathing. Applying force fluctuations of physiological amplitude and timing to contracted ASM causes the contracted muscle to relengthen substantially (9–10, 13, 29). The amount of relengthening can be altered by intervention of specific signaling pathways governing muscle activation. Inhibition of p38 MAPK and the upstream activator of ERK1/2 (i.e., MEK) potentiates relengthening of maximally activated ASM (8, 28); this potentiation appears to be mediated by regulation of phosphorylation of the 27-kDa heat shock protein (HSP27) (27) and h-caldesmon, a protein known to cross-link the thin and thick filaments (8). It is not clear how phosphorylation of HSP27 and h-caldesmon is linked to force-fluctuation-induced relengthening, but it is speculated that the phosphorylation may affect the thin filament length, leading to shorter contractile unit length and reduced actin-myosin-actin connectivity within the contractile units, which in turn translates into a reduced overlap between the contractile filaments during the length oscillation (caused by the force fluctuation) (30).

In force-fluctuation-induced ASM relengthening, some degree of plastic deformation can be observed. In going from small- to large-amplitude oscillation, the amount of relengthening increases; however, when returning from a large-amplitude oscillation to oscillation at a smaller amplitude, the amount of relengthening is not accordingly reduced, but remains at about the same level as that observed during the large-amplitude oscillation (13, 28). It appears that length adaptation and structural malleability may play a role in this nonreversibility.

Cytoskeletal reinforcement.

In response to the application of a localized physical force, as by an attached microbead, there occurs rapid actin polymerization and increased focal adhesion assembly, resulting in increases in cytoskeletal stiffness and traction forces (41, 61–62). This phenomenon is called cytoskeletal reinforcement. But if left unopposed, reinforcement would progressively impede cell stretch, and organ stretch, and thus could become a self-defeating strategy of cytoskeletal mechanoprotection. To maintain a homeostatic balance, an opposing phenomenon is required. Indeed, fluidization is now seen as being reinforcement's opposite (5, 23, 57). In response to a uniform biaxial or uniaxial cell stretch, the cytoskeleton exhibits a prompt decrease of stiffness and increase in macromolecular mobility (23, 57); an example is shown in Fig. 2. In response to stretch, therefore, the cell might either reinforce, a bracing-type of physiological response, or fluidize, a stress-relieving physiological response.

Fig. 2.

Cell mapping rheometry of human airway smooth muscle cell. A: traction map before cell stretch. B: traction map immediately after an imposed homogeneous biaxial stretch of a 4-s stretch-unstretch maneuver with a peak strain amplitude of 10%. The cell tractions are markedly ablated. C: traction map at 1,000 s following stress cessation. Tractions have largely recovered to the prestretch value. D: traction field can be used to compute the contractile moment, T, corresponding to an equivalent force dipole. At the earliest measurable time point following stretch (b), the contractile moment was significantly reduced to 20% of its baseline value (a) followed by a slow recovery (c). Reproduced from Ref. 23 under its Creative Commons Attribution License.

Cytoskeletal fluidization and resolidification.

The fluidization response is prompt and is mediated by the effects of physical forces acting directly on a material, the cytoskeleton, that is innately fragile (3–5, 7, 11, 12, 23, 31, 51, 57, 58, 71). It remains unclear, however, what the underlying mechanism might be. Perturbed myosin binding is certainly a large contributor; physical forces can pull myosin heads away from actin filaments, but other weak cross-linking bonds can be disrupted as well (24, 44, 63, 65). Depolymerization of F-actin in response to stretch has also been known for a long time (37), but based on existing data, that depolymerization process was thought to be too slow to account for prompt cytoskeletal fluidization. Recent data, however, have implicated stretch-induced actin depolymerization (5); evanescence of actin filaments is described in greater detail below.

Strain-rate invariance.

To the extent that the ASM cell is seen as being a highly malleable material, it shares many features with soft inert materials such as foams, clays, pastes, colloids, and emulsions, all of which, like ASM, show relaxation dynamics that are not tied to any internal scale of time (40, 48, 49). For oscillatory cell stretching, strain-rate amplitude is proportional to the product of the length change and the frequency; for a muscle of unit length, the strain-rate amplitude corresponds to the peak velocity in the oscillatory cycle. When viewed as a function of the strain-rate amplitude alone, length adaptation, on the one hand, and actomyosin bridge dynamics, on the other, can be unified (34). To explain this unification, Oliver et al. (34) pictured molecular rearrangements within the cytoskeleton as being governed by long-lived microconfigurations in which stress-bearing molecules become trapped. For small strain rates, the relaxation time must remain close to its natural (slow) time scale, whereas for larger strain rates another time scale enters into the problem, namely, the strain rate itself (66).

CHANGES IN CELLULAR CONSTITUENTS AND THEIR ORGANIZATION DUE TO MECHANICAL PERTURBATION

A common requirement for the emergence of the above-described smooth muscle behaviors is structural malleability of the cell. This property probably originates from multiple levels of the structural organization of the muscle cell that include weak interactions among the structural elements of the cytoskeleton and between actin and myosin within the confine of the contractile unit, linkages between structural elements that are readily modifiable, and lability of the structural elements themselves. Some of these structural remodelings may be mediated by chemical signaling triggered by mechanosensors on the cell membrane; others may be a direct result of mechanical perturbation.

Evanescence of myosin and actin filaments.

The lengths of thick and thin filaments in striated muscle are fixed, and they do not change under physiological conditions (50). In smooth muscle, there is no consensus regarding the lengths of the thick and thin filaments; there is strong evidence suggesting that the filament lengths are not constant (45).

Myosin monomers of smooth muscle are able to self-assemble into filaments. It has been shown that the filaments are not stable when the regulatory light chain of the myosin is not phosphorylated, while filaments of myosin with the light chain phosphorylated are much more stable (42, 52). It is further shown that a conformational change in the phosphorylated vs. nonphosphorylated myosin is likely responsible for the filament stability (6, 35, 47, 59, 60). Direct test of the physical integrity of ASM myosin filaments confirms that phosphorylated myosin monomers are able to form filaments that are much more resistant to ultrasonic agitation in solution (20). The above cited studies were all carried out in solution. In intact ASM cells it has been shown later that the light chain phosphorylation also regulates thick filament lability (39).

Lability of thick filaments in intact ASM can be demonstrated in at least two ways. An increase in thick filament mass has been shown to be associated with muscle activation (16, 17, 25); mechanical agitation in the form of length oscillation applied to relaxed ASM has also been shown to transiently reduce the thick filament mass (26). The change in filament mass is completely reversible if the muscle is allowed to recover under static conditions. Interestingly, change in isometric force due to length oscillation parallels that of thick filament mass (26). These findings suggest that thick filaments disassemble in response to externally applied strain and reassemble in the absence of mechanical disturbance. The evanescence of thick filaments is likely one of the important intrinsic properties of ASM that gives rise to the phenomenon of length adaptation (25).

The thin filaments in smooth muscle, like the thick filaments, are not static structures resistant to remodeling. Actin polymerization has been shown to occur with activation of ASM (17, 19, 21, 33, 56), and the polymerization is critical for normal force development (33). Polymerization of actin filaments in ASM is extensively regulated; increasing details of the signaling pathway are being revealed mainly due to the work being done in the Gunst laboratory (15). It appears that regulation of actin dynamics in smooth muscle shares some similarity with that in migrating cells. It has been demonstrated that the actin nucleation promoting factor, N-WASp, is involved in the activation of the actin nucleation protein, Arp2/3, that in turn leads to actin polymerization (69). When the intracellular process of actin polymerization is disrupted, tension development in the muscle in response to stimulation is invariably attenuated (54, 55, 69, 70). Involvement of integrin-linked kinase in actin polymerization (70) suggests that mechanical stress or strain could be an important trigger for actin network remodeling.

Plasticity of the dense-body structure.

Another structural element in ASM that exhibits malleability is the dense body (DB) cable, a longitudinal aggregate of dense bodies flanked by intermediate filaments. It has been found in ASM that there are numerous DB cables that run in parallel with the contractile filaments within any cell segment (68). These cables appear to be able to support passive tension (in the relaxed muscle) because stretching the muscle cells straightens the cables (68). The DB cables behave like ordinary cables, able to support tensile stress when the muscle cell is suddenly stretched, and undergo “stress relaxation” when the cell is maintained at the stretched length. However, they do not behave like ordinary cables when the muscle cell is shortened. These cables do not buckle but instead are able to adjust their length (within seconds) and remain more or less straight even when the cell length is reduced by half (68). It appears that the DB cables are plastic structures able to alter their lengths according to cell length. This property may be responsible for the constant stiffness in the relaxed ASM observed in a wide range of adapted cell lengths (2, 68).

Vimentin is one of the major proteins that make up intermediate filaments. Regulation of the length of DB cables is likely linked to regulation of the length of the intermediate filaments. It has been shown that phosphorylation of vimentin occurs during smooth muscle activation and that phosphorylation of vimentin leads to disassembly of intermediate filaments (32, 53, 64). This may be the molecular mechanism by which the length of DB cables is regulated. That is, intermediate filaments disassemble on activation, and when the muscle settles to its final resting length after relaxation and dephosphorylation of vimentin, reassembly of intermediate filaments at a different length occurs.

Dynamics in thin filament attachment to dense plaques.

In skeletal muscle, a single cell spans from tendon to tendon, and the cell-tendon connection is a permanent structure in that its structural organization does not change with the contraction-relaxation cycle. In cardiac muscle, individual myocytes are connected to one another through intercalated discs which are also permanent structures. In smooth muscle, there is evidence that the dense plaques that couple adjacent cells mechanically may not be permanent structures (15). It has been demonstrated in ASM that recruitment of structural and signaling proteins to the cell cortex is associated with muscle stimulation, and that this recruitment stimulates association of adhesion proteins to β-integrins (22, 36, 67, 69–70) presumably at the sites of dense plaques. Some of the important events that occur at the adhesion sites during activation are polymerization of actin filaments and attachment of the filaments to anchoring proteins, which also provide linkages to integrins and the extracellular matrix.

While the permanent mechanical connection adopted by striated muscle may be energetically more efficient, it does not provide malleability required by smooth muscle. For length adaptation to occur, complex mechanisms regulating dense-plaque function must be in place to ensure that the cytoskeleton and the mechanical couplings are rigid enough for force generated by the muscle to be transmitted to the outside world, and yet flexible enough that the structures can be modified to accommodate large change in cell geometry.

UNDERSTANDING SMOOTH MUSCLE FUNCTION IN RELATION TO STRUCTURAL MALLEABILITY

It is obvious that many of the emergent smooth muscle properties stem from structural malleability of the cell. How structural properties determine functionality in smooth muscle is, for the most part, still an unanswered question. To apply the systems biology approach in understanding smooth muscle function, we need to integrate data at all levels, from protein-protein interaction to the manifested mechanical properties. Recent development in smooth muscle research, especially in ASM research, has switched our focus on regulation of actomyosin interaction to a broader focus that includes regulation of structural changes within the muscle cells during contraction and relaxation. Given that structural alterations in cytoskeleton and contractile apparatus can affect smooth muscle force and stiffness development (2, 26, 33, 38, 46, 67, 68–70), velocity of shortening, (25, 38, 46) power output (25), and relengthening during contraction (8, 13, 27–28), we can no longer interpret mechanical data from smooth muscle solely based on models of actomyosin interaction.

Smooth muscle shares with nonmuscle motile cells many features in terms of their activation pathways. One can speculate that the reason behind this is the common need for flexibility in the subcellular structures. In order to maintain structural malleability, smooth muscle appears to have taken a separate path in evolution from that taken by striated muscle. The latter appears to have streamlined the activation pathway for efficiency, because it does not need structural malleability in order to operate over a large length range. This insight is probably useful in guiding our research approach and puts less emphasis on the similarities between smooth and striated muscles, and more on the similarities between smooth muscle and motile nonmuscle cells. The rich literature dealing with regulation of motility in nonmuscle cells has been, and will likely continue to be, very relevant to smooth muscle research (15).

Although progress has been made, there are still many challenges in our attempt to understand emergent smooth muscle function based on specific molecular interactions and structural modifications. A major one is the lack of sufficient ultrastructural data, specifically those dealing with changes at the ultrastructural level due to mechanical or chemical interventions. Therefore there are still many gaps in the links between specific molecular mechanisms and the emergent behaviors. There is also a lack of sufficient information regarding the interactions among the various molecular processes that lead to structural modification and the emergent properties. Nevertheless, by overcoming these challenges we will be rewarded with opportunities to identify drug targets that can change the outcome of molecular interactions, thus providing guidance for therapeutic approaches in treating many smooth muscle-related diseases.

GRANTS

This work was supported by the Canadian Institutes of Health Research (MOP-13271, MOP-37924) and the National Institutes of Health (HL-084224).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).