INTRODUCTION

TOP ABSTRACT INTRODUCTION DEFINITION OF CONTRACTILITY MECHANICS OF THE ASTHMATIC'S THE MECHANICAL PROPERTIES OF PHENOTYPIC HETEROGENEITY OF CONCLUSIONS REFERENCES

It speaks volumes for how far airway smooth muscle (ASM) research lags behind skeletal muscle research when questions about the contractile apparatus of ASM and its function can only be answered in terms of the properties of strips of ASM. In terms of subcellular and molecular level characterization, much remains to be done. Part of the problem is that no organized sarcomeric structure has been described for smooth muscle to which to relate its function. The structure-function relationship for smooth muscle still awaits elucidation.

Until 20 years ago, although myosin protein was known to be present in smooth muscle, thick filaments had not been seen. This was attributed to depolymerization of the filament during preparation of the tissue for microscopy. In 1984, using refined techniques, Somlyo and colleagues (1) published electron micrographs of longitudinal sections of smooth muscle. These showed clearly defined thick filaments running between thin filaments with interlinking crossbridges. Dense bodies were also identified and studied by Fay and Fogarty (2) with the aid of -actinin antibodies. They showed they are symmetrically aligned across the cytoplasm, serve as attachment sites for actin filaments, and resembled the z-discs of striated muscle. Somlyo also showed arching fibers between adjacent dense bodies. Immunologically they were identified as desmin. This was akin to what was seen in skeletal muscle where desmin interconnected adjacent z-discs and thus stabilized sarcomeres. Using so-called "decoration" techniques in which isolated myosin heads were made to interact with actin filaments, Somlyo demonstrated polar reversal of myosin binding sites on actin subunits on either side of a dense body. This further supported the idea that dense bodies are z-disc analogues. Evidence is thus accruing to show that quasi-sarcomeric structures do exist in smooth muscle and some progress is being made.

To put the ASM problem in its proper perspective, one should point out that research on skeletal muscle crossbridges is currently centered on definition of the structures of the crossbridge, namely the head of the myosin (3) and actin molecule (4) at atomic level, and the development of assays to delineate mechanics and biochemistry at the single molecule level (5). Rayment and colleagues' work (3) on the crystal structure of the myosin head sets out the rules for the function of the molecule. Current experiments with reflectance fluorescence microscopy are providing direct evidence for a relationship between mechanical events and nucleotide turnover at the level of a single myosin head (8).

Perhaps the most important part of the myosin head is its domain 20 (20 kilodaltons [kD]), which likely acts as a lever to amplify small 0.5-nm conformational shifts observed in the crystal structures of the carboxy termini of the myosin heads into macroscopic steps of the order of 10 nm of the lever for each round of adenosine triphosphate (ATP) hydrolysis.

To show that smooth muscle research is not lagging that far behind, it must be pointed out that although the crucial bent conformation of the 20 kD lever during adenosine diphosphate (ADP) binding has not been seen when the skeletal myosin head binds to actin, it has been reported for smooth muscle (9).

To evaluate the current state of ASM research, it must be recognized that studies at the organ level or of ASM strips cannot afford insights into subcellular mechanisms. This is because of the demonstrated structural, biophysical, and biochemical phenotypic heterogeneity of the constituent cells. Our own experience with phenotypic heterogeneity of airway smooth muscle will be dealt with below. Because of this heterogeneity, mechanistic studies, both biophysical and biochemical, must be conducted at a single cell level.

Though deficits in knowledge of structure and function for ASM exist, windows of opportunity are opening that may allow smooth muscle research to be conducted at the crossbridge level. Lauzon and coworkers (10), for example, have successfully applied the double laser trap methodology to motility assay research in smooth muscle.

Why all this preoccupation with molecular mechanisms for ASM contraction? The answer is that, in disease, where structure and function are disordered, the ultimate cause is an altered molecule and the necessary therapy could well be that engineered with the techniques of molecular biology.

	DEFINITION OF CONTRACTILITY

TOP ABSTRACT INTRODUCTION DEFINITION OF CONTRACTILITY MECHANICS OF THE ASTHMATIC'S THE MECHANICAL PROPERTIES OF PHENOTYPIC HETEROGENEITY OF CONCLUSIONS REFERENCES

The definition should apply to both intact muscle and to interacting molecules of the muscle's molecular motor, and encompass all the functions the muscle carries out. These functions are to stiffen and thus bear steady-state loads, which is best assessed by measuring maximum developed isometric force (P_o); to shorten and thus effect airway narrowing (this should be load-independent, i.e., zero load-shortening [L_max]); to achieve shortening at maximum velocity (V_o) (this should also be load-independent); and to achieve the above in the presence of an energy utilization rate, which is at a steady maximum. This is also regarded as holding the effect of time (t) on contraction constant. It is achieved by using a supramaximal tetanic electrical stimulus. In short, contractility is defined as the instantaneous relationship between force, velocity, length, and time. Simultaneous measurement of three of these without determining the status of the fourth cannot be accepted as measurement of contractility.

	MECHANICS OF THE ASTHMATIC'S AIRWAYS

TOP ABSTRACT INTRODUCTION DEFINITION OF CONTRACTILITY MECHANICS OF THE ASTHMATIC'S THE MECHANICAL PROPERTIES OF PHENOTYPIC HETEROGENEITY OF CONCLUSIONS REFERENCES

With respect to asthma research, considerations additional to these described above must be dealt with. These resolve into four questions:

1. Is the contractility of bronchial smooth muscle (BSM) increased in the airways of patients with asthma? The consensus is that the muscle itself evinces normal contractility but is made to contract more forcibly by increased amounts of released agonists. The reason this question is so far unanswered is that it is practically impossible to conduct the requisite experiments in statistically significant numbers on tissues obtained from living subjects with asthma.
2. What is the role of ASM in regulating distribution of ventilation? This issue has not been resolved although a host of speculations exist. Otis (11), in an incisive review, pointed out the lack of conclusive evidence. The speculations are that airway narrowing, by reducing anatomical dead space, would improve alveolar ventilation. This, of course, could only occur if respiratory frequency was not too greatly increased. Regional constriction is said to regulate local distribution of ventilation. Finally, the squeezing action of peristaltic activity of BSM could help eliminate mucus.
3. What are the important factors regulating bronchoconstriction? The recent observation that the bulk of ASM shortening occurring in the first 3 s of the muscle's contraction time (10 s) needs to be considered (12). We have found that 90% of total shortening is completed within 3 s. Since any further shortening is negligible, studies at steady state are misleading. Furthermore, as shortening is to be completed in 3 s, clearly V_o is an important limiting factor, much as is the case for cardiac muscle, which can only respond by twitches. The factors regulating V_o are the activity of smooth muscle myosin light chain kinase, which activates actin-activated myosin Mg²⁺-ATPase activity, and the compliance of the ASM cells' so-called internal resistor (13).
4. What is the load on the shortening muscle in vivo? Under normal circumstances it is felt to be the elastic recoil of the attached lung parenchyma to the airway wall (14); this has been referred to as the interdependence effect. However, on the basis of video-imaging morphometry of isolated airways, a minority view (15, 16) suggests that this type of radial traction by the parenchyma is not responsible for the load on the muscle, which, instead, is the loose connective tissue attaching the external aspect of the ASM layer to the internal perichondrium of superjacent cartilagenous plaques (17). This load could be quasi-isotonic but it is more likely elastic.

One very important mechanism recently proposed by Fredberg and associates (18) suggests that a steady-state force balance does not exist between parenchyma and ASM but is replaced by a dynamic equilibrium state between the cycling rate of ventilation-dependent stretching of the parenchyma and the cycling rate of ASM crossbridges.

	THE MECHANICAL PROPERTIES OF AIRWAY SMOOTH MUSCLE

TOP ABSTRACT INTRODUCTION DEFINITION OF CONTRACTILITY MECHANICS OF THE ASTHMATIC'S THE MECHANICAL PROPERTIES OF PHENOTYPIC HETEROGENEITY OF CONCLUSIONS REFERENCES

In general, studies of the contraction mechanisms of ASM (length-tension [LT] relationships) should be conducted at a single crossbridge. Preliminary work in this direction has commenced and will be dealt with below. First, more conventional studies at muscle strip level will be discussed.

Tracheal Smooth Muscle

Figure 1 shows LT curves elicited from tetanized canine tracheal smooth muscle (TSM) (16, 19). They resemble in qualitative terms those for skeletal muscle (20) and, in more quantitative terms, those reported for other smooth muscle (21). The major characteristics of the tracheal smooth muscle curves are:

1. The considerable compliance of the resting muscle, which displays very little resting tone. This is quite unlike that for human BSM, where resting tone (force) at optimal length (l_o) is almost equal to maximum active isometric force (P_o) developed by the muscle (22). The physiologic implication is that while active force development would be required to shorten the former, less would be required for the latter where passive recoil forces are available to aid shortening.
2. The value of P_o is equivalent to that of skeletal muscle and of all smooth muscles in general. This is remarkable when one considers that smooth muscle has only one-fifth the myosin content of skeletal muscle. Part of the explanation is that the myosin filament in smooth muscle is 2 µm long while that in skeletal is 1.65 µm. Thus, the former possesses more crossbridges per half sarcomere and can exert more force. Another part of the explanation stems from the fact that ASM shortens with one-hundredth the velocity of skeletal muscle. The contact, or duty time, between the actin and myosin molecules would be much longer and thus allow generation of more force.
3. The ability of the isolated lightly loaded TSM to shorten by almost 90% of its initial length (l_o). This again is considerably greater than the 35% of skeletal (20) muscle and could contribute to severe bronchoconstriction. In vivo and in situ where muscle attachments are intact, and auxotonic loading is operative, the shortening is much less.
4. The marked increase in stiffness of the active muscle, which can be gauged by comparing the slopes of the resting tension and total tension curves. Perhaps the most important conclusion to be drawn from LT curves is that the amount of shortening that can be estimated from such curves is critically dependent upon the mode of contraction. This is seen in muscle strips contracting isometrically, or in after-loaded or free-loaded shortening mode (23). Whether the loading is isotonic or auxotonic also plays an important role. This is discussed later.

View larger version (31K): [in this window] [in a new window]   Figure 1.   Length-tension curves (± SE) of canine tracheal smooth muscle. Lmax is the length at which maximum isometric active tension (Po) develops. Reprinted by permission from Reference 16.

Bronchial Smooth Muscle

LT curves for BSM (12, 24, 25) were obtained from third lobar generation BSM. They strongly resemble those of TSM. An apparent striking difference is in the value of P_o, when P_o, as is the convention, is normalized with respect to tissue cross-sectional area. However, while 75% of the cross-sectional area of TSM is muscle, the content in BSM is only 20% (24). Once the normalization is made with respect to muscle cross-sectional area alone, the difference becomes much less.

Before discussing what is the true after-load on shortening ASM, cognizance must be taken of an important report that indicates our interpretation of the shape of the LT curve must be significantly altered.

The New Length-Tension Curve

Pratusevich and associates (26) showed that when a small segment of a strip of TSM is held constant, and, when at any given steady-state length, the muscle is tetanically stimulated five times at 5-min intervals, active tension does not decrease at lengths ranging from 0.5 l_o to 1.5 l_o but becomes length- independent. This remarkable finding forces us to reevaluate our ideas of the meaning of the length-tension curve. If the LT curve is elicited in the conventional way, viz., single tetanic stimulations of a strip of muscle, a conventional Frank-Starling type of active tension curve is obtained. Since this represents the way in which these relationships are perhaps evoked in vivo in the airway wall, the Frank-Starling response is the physiologic response that results from interactions between active cells and passive connections. However, subcellular and molecular mechanisms can best be studied at fixed segment length, and mechanistic studies must be conducted in this mode.

Ford's explanation for this unique behavior is that new filaments of myosin are synthesized from monomers during the repeated stimulations cited above (26). These filaments are laid down in series, thus accounting for the length-independence of force development. The in-series filaments would also account for the increased compliance and maximum velocity of shortening (P_o) that they found. Evidence in support of new filament formation is reported by Gillis and colleagues (27), who observed changes in birefringence in a contracting smooth muscle, which indicated new structure development. Onishi and Wakabayashi (28) and Trybus and Lowey (29) have also obtained results to indicate new filament formation. Seow has obtained preliminary evidence (unpublished results) based on reduced nicotinamide adenine dinucleotide (NADH) fluorescence measurements of crossbridge energetics in TSM that supports the idea that new filament formation is indeed occurring during repeated stimulation of a fixed segment of muscle.

Figure 2 shows LT curves elicited by us from canine TSM. The curve on the right shows a steep LT curve that displays the usual Frank-Starling shape. This curve was elicited in the conventional way from an strip stimulated once tetanically. The curve in the middle was elicited from a monitored segment of the muscle that was stimulated just once. The left curve was elicited from a fixed segment stimulated four times. The independence of P_o from length is seen over a range extending from 0.5 l_o to 1.0 l_o which resembles that reported by Pratusevich and colleagues (26).

View larger version (15K): [in this window] [in a new window]   Figure 2.   Segmental length-tension (LT) curves elicited from canine tracheal smooth muscle. A selected segment (100 µm long) in the middle of the muscle strip was held constant electronically throughout contraction. At any given length the muscle was stimulated tetanically for 10 s at 5-min intervals. It was stimulated three times in this mode. The right curve is the conventional LT curve in which the muscle is stimulated once at the selected length. The middle curve was elicited from a muscle segment but with only one stimulation. The left curve is from a segment stimulated twice at each length. Note that no further tension was developed with subsequent stimulation. Reprinted by permission from Reference 16.

The Special Nature of Isotonic Shortening in Airway Smooth Muscle

Delineation of shortening behavior. Isotonic shortening in ASM is unlike that of striated muscle. Figure 3 shows an isotonic shortening versus time curve for a lightly loaded strip of BSM (12). The duration of the supramaximal electrical stimulus is depicted above the shortening trace. Steady-state maximum shortening occurs at 8.5 s after onset of shortening. However, it is clear that 75% of the shortening is complete 1.5 s after stimulus onset, while 90% is complete within 3 s; thereafter, for all practical purposes, shortening is independent of length. Thus, with respect to regulation of airway resistance, the first 1.5-3.0 s are all-important.

View larger version (16K): [in this window] [in a new window]   Figure 3.   Record of shortening versus time elicited from a muscle shortening isotonically with a light pre-load; 75% of total shortening is complete within 2 s of onset of stimulation. Ninety percent is complete within 3 s. Reprinted by permission from Reference 12.

Another way of showing the importance of the early shortening phase is to plot the difference in isotonic shortening between ragweed pollen-sensitized canine BSM and the same generation BSM from a litter-mate control subject. The increased shortening difference of the sensitized muscle is complete at 2.5 s, after which no further increase in shortening difference occurs. Thus, in allergic bronchospasm, and likely in asthmatic bronchospasm, it is the early phase of contraction that is important.

The role of crossbridges. Figure 4 shows a plot of lightly loaded isotonic shortening versus time for ASM (16). At differing times in successive sweeps the load was abruptly (within 0.5 ms) changed to zero, generating the so-called zero load clamp. For each clamp a rapid transient, representing the elastic recoil of the muscle's series elastic component, is seen. This is followed by a slow transient that represents the shortening of the muscle's contractile element. The maximum slope of this transient represents the maximum velocity of shortening (V_o). These V_os diminish with time.

View larger version (18K): [in this window] [in a new window]   Figure 4.   Zero-load clamps applied to a lightly loaded isotonically contracting trachealis. Reprinted by permission from Reference 16.

When zero load clamps are applied during the course of an isometric contraction, a time-dependent decrease in V_o is still evident, but it is much less than that seen in Figure 4.

Figure 5 shows plots of V_o versus time for the two modes of contraction. Maximum V_o values are obtained at the same instant, i.e., 1.5 s in both modes. Thereafter, while a 10-fold drop in V_o occurs when quick releases in isotonic shortening mode are employed, only a 0.8-fold decrease is seen when releases are made in the course of an isometric contraction. These drops can be explained by the development of latchbridges (30) or slowly cycling bridges (31) in the latter part of contraction. The drop in canine TSM in isometric mode is much less than that reported for hog carotid smooth muscle by Dillon and coworkers (30). We have no explanation for this difference other than the differences in animal species and tissues. The important conclusion is that the first 2-3 s of shortening are carried out by normally cycling (myosin light chain [MLC]20 phosphorylated) crossbridges in muscle strips whose shortening velocity is 0.41 l_o/s (32). It is in this period that 75% of L_max and maximum V_o develop.

View larger version (14K): [in this window] [in a new window]   Figure 5.   Plot of Vo versus time derived from Figure 4. Reprinted by permission from Reference 16.

As stated before, it is felt L_max should not be time-limited, and given sufficient time the ASM should achieve maximum shortening. However, in tetanized ASM, since 90% of the L_max is completed within 3 s, velocity of shortening (V_o) becomes an important determinant of L_max, just as it does for cardiac muscle. It follows that smooth muscle type myosin light chain kinase (sm-MLCK) activity is paramount in determining L_max. The ultimate factor is, of course, the actin-activated myosin Mg²⁺ ATPase activity of the muscle. Since the relation between this and phosphorylation of MLC20 is linear, the latter is employed as a convenient index of the former. The other parameter limiting L_max is the compliance of the muscle's internal resistor. Its properties have been delineated by Stephens and colleagues (13) for ASM. Their study showed additionally that the compliance of the internal resistor of sensitized canine ASM was greater than that of the litter-mate control subject. This could account at least partly for the increased L_max of the ASM (13, 16). At the extracellular level this resistor could be stemming from the extracellular matrix, and at the intracellular level from the cytoskeleton (-actin, desmin, vimentin, talin, vinculin, metavinculin, and paxillin) or perhaps the latch bridges. In connection with the activities of the latch bridges, important modulators are calponin (34) and caldesmon, with the former being the more important.

At steady state (8.5 s) the muscle is dedicated to developing force, at which time the important regulator is dephosphorylation of the 20-kD myosin light chain brought about by myosin light chain phosphatase.

The role of auxotonic shortening. So far it has been shown that time-dependence is important to ASM shortening. It appears that the nature and history of loading of the muscle is also important. Figure 6 shows length-tension curves (12) elicited from TSM using an electronic auxotonic loader. The loader applies isotonic, exponential, sigmoidal, elastic, and logarithmic loads. In the figure the final load achieved was the same for all the differing loading modes, but the shortening achieved was quite different. The explanation lies in the fact that since maximum shortening is achieved within the first 2- 3 s, when normally cycling crossbridges are active, loading in this interval will maximally regulate shortening. Thus, in logarithmic mode since the loading is highest, shortening will be the least. Clearly the nature of in vivo loading is of the greatest significance in determining shortening.

View larger version (15K): [in this window] [in a new window]   Figure 6.   Auxotonic shortening-force plot under different loading modes. Length-tension curves of trachealis shortening under logarithmic, linear, sigmoidal, and exponential loading modes (reading from left to right). It is clear that maximum shortening differs with loading mode. The least shortening is seen where heaviest loads (logarithmic) are applied early in the contraction. Reprinted by permission from Reference 16.

The nature of in vivo loading of airway smooth muscle. Several sources of loading have been identified and include, for example, the epithelial layer, blood vessels in the airway walls, intramural exudates and transudates, mural inflammatory cells, connective tissue linking the external ASM cell layer to the perichondrium of the superjacent cartilaginous plaques, and the lung parenchyma. The last mentioned is said to be the most important member of the list. Ding and associates (14) showed that the load on BSM varied as a function of lung volume. This suggested the lung parenchyma attached to the airways exerted elastic traction, producing an interdependence effect. In acute asthma, the transient increase in compliance of the lung parenchyma would reduce the load on the BSM, thus resulting in increased shortening of the muscle and narrowing of the airways.

Recent reports by Mitchell and Sparrow (15) from Australia and Stephens and colleagues (16) indicate a different viewpoint, although it is in the minority. Study of contraction of a cylindrical segment of isolated airway with the aid of video imaging and laser endobronchoscopy reveals that though considerable reduction of the limenal cross-sectional area occurs, the external cross-sectional area of the airway does not. Microscopy also reveals that the space between muscle layer and cartilage increases. One may infer from this that the true load on the muscle is the rather loose connective tissue between muscle and cartilage. This load is likely to be elastic in nature, though in view of its paucity and slackness could be quasi-isotonic.

Of all the factors mentioned above, it seems the one most important responsible for bronchoconstriction is ASM shortening, with the others making minor contributions.

	PHENOTYPIC HETEROGENEITY OF AIRWAY SMOOTH MUSCLE

TOP ABSTRACT INTRODUCTION DEFINITION OF CONTRACTILITY MECHANICS OF THE ASTHMATIC'S THE MECHANICAL PROPERTIES OF PHENOTYPIC HETEROGENEITY OF CONCLUSIONS REFERENCES

Following closely upon the heels of vascular smooth muscle research wherein structural, functional, and biochemical heterogeneity was first reported (33), similar heterogeneity has been reported in ASM (35, 36). The implications are that in elucidating molecular mechanisms of contraction in normal and diseased airway smooth muscle, studies will have to be conducted on single cells since the assumption of homogeneity of cells in a strip of smooth muscle is no longer tenable.

Heterogeneity at Tissue Level

We have reported (35) that the mechanical properties of canine main pulmonary arterial strips are quantitatively different from those of tracheal smooth muscle. The capacity (L_max) and velocity of shortening (V_o) were considerably less in the pulmonary arterial strip. Marked changes in contractile regulatory and cytoskeletal proteins were also present.

Heterogeneity at Cellular Level

Cell length. In studies conducted on enzymatically separated single smooth muscle cells from the canine TSM, we found both structural and contractile heterogeneity.

Figure 7 shows cells of uniform length. Ninety five percent were fully viable as judged by dye exclusion and by contractile response to electrical stimulation. Figure 8 shows another sample of TSM cells. Two types of cells are seen: one has the same dimensions as in Figure 7 while the others are much longer. Measurements of 300 cells per experiment were conducted in six experiments. The cells were studied after complete pharmacologic relaxation or in zero calcium solution. There seemed to be no evidence of contracture.

View larger version (109K): [in this window] [in a new window]   Figure 7.   Freshly isolated cells from canine tracheal smooth muscle seen under phase contrast. The refractile cell margins indicate viable cells. The cells are fairly homogeneous with a mean length of about 100 µm.

View larger version (89K): [in this window] [in a new window]   Figure 8.   Freshly isolated cells from canine bronchial smooth muscle and seen under phase contrast. Both short (length, 110 µm) and long cells (200 µm) are seen.

Frequency distribution of TSM cells lengths was examined. Two distributions were discerned. In one the mean length was 113 µm ± 2 µm (SE). These constituted 83.7% ± 3.6% (SE) of the total population. In the other mean length was 191 µm ± 9 µm (SE). They constituted 16.3% ± 3.6% (SE) of the total. A statistical technique called the method of mixtures (37) was made available to us by Dr. MacDonald of McMaster University and was used to identify the different populations. The shorter cells were arbitrarily termed Type I. The longer cells were labeled Type II.

Similar studies were made on cells obtained from bronchial smooth muscle (BSMC). They showed the same structural characteristics as those of tracheal smooth muscle cells (TSMC), but their frequency distributions were different. The smaller cells, Type I, were 105 µm ± 3 µm (SE) and they constituted 42.1% ± 3.9% (SE) of the total, while the longer, Type II, were 200 µm ± 4 µm (SE) and made up 57.9% ± 3.9% (SE). Thus, structural heterogeneity is evident. Interestingly, in BSMC obtained from dogs sensitized to ragweed pollen, the length distributions were almost the same. Type I cells were 99 µm ± 7 µm (SE) in length while Type II were 200 µm ± 4 µm (SE). However, the frequency distributions differed between control and sensitized TSMC. They were 15.4% ± 4.1% (SE) for Type I and 84.6% ± 4.1% (SE) for Type II in sensitized TSMC.

Cell contractility. Shortening capacity (L_max) and V_o were studied concomitantly with the length studies. The cells were free-floating on the slide, and shortening was measured using video imaging. They could be stimulated to contract using acetylcholine (10⁵ M), KCl (80 mM), or histamine (10⁴ M). On washing out the stimulants, the cells relaxed fully. The cells could also be made to shorten reversibly with a single electrical impulse delivered by a platinum electrode.

Figure 9 shows a Type I cell. The tip of the electrode was 5 µm in diameter, which was equivalent to the thickness of the Type I cell shown. Because of the angle of the lens selected to optimize the image of the cell, the image of the electrode itself was distorted. Figure 10 shows the same cell after stimulation. About 25% shortening is seen. Newly formed small vesicles are visible on the sarcolemma at the lower end of the cell. One subtype of Type II cells shows 20% shortening of the cell. Another type of Type II cell shows a shortening of about 65%.

View larger version (87K): [in this window] [in a new window]   Figure 9.   Video image of a single normal canine tracheal smooth muscle cell. The black object is a platinum electrode seen obliquely and hence distorted. The tip measured 5 µm, which would be equivalent to the width of the cells. The cell shown here has been arbitrarily labeled Type I.

View larger version (87K): [in this window] [in a new window]   Figure 10.   The same cell shown in Figure 9 after stimulation by a single electrical pulse. A 25% shortening of the cell is evident.

Populations of such cells were studied. Type II cells were divided into IIA, which shortened by 20%, and IIB, which shortened by 65%. Using video imaging, frame grabbing, data digitization, and custom-designed software, we obtained records of cell shortening versus time. From these records both L_max and V_o for the various cell types could be derived. Mean results from a series of experiments confirmed the existence of three types of cells: Type I, which are short in length and shorten by a small extent; Type II, which are long but shorten to a similar extent as Type I; and Type IIB, which are long and shorten considerably more (65%). Phenotypic heterogeneity therefore becomes extremely important in determining function. Similar results were found with respect to V_o. The control of L_max by V_o has been referred to already. With respect to data obtained from airway smooth muscle cells (ASMC) of sensitized and litter-mate control dogs, we found that the distribution of cell types is not altered by sensitization. This point is of significance in view of the fact that sensitization is started on the day of birth and the actual studies are made at 5 or 6 mo of age. Regardless of the cell type, sensitization increases shortening capacity in all.

Alterations of Airway Smooth Muscle Contractile Phenotype

We have found two such alterations, one in situ and the other in culture. We have reported (16) significant change in contractile phenotype as a result of sensitization. At levels corresponding to the preload, which sets the muscle at strip at l_o, the sensitized muscle shortens by an additional 15% vis-á-vis the control. This corresponds to a 50% increase in lumenal resistance. Force-velocity curves show that while the maximum isometric force is unchanged, V_o is increased with sensitization. We have also shown that the increase in L_max and V_o is the result of cycling activity up to 3 s following the stimulus. Thereafter there is no further increase, i.e., it is only the normally cycling crossbridge that shows increased contractility. We have also shown identical results in BSM strips. The sensitized TSM and BSM showed increased actin-activated myosin Mg²⁺-ATPase activity, which would account for the increased V_o shown by these muscles (16). Our studies also show that this is not due to any change in the contents of myosin heavy chains. Contents of smooth muscle myosin heavy chain 1 (SM-1), heavy chain 2 (SM-2), and SM-B, a recently discovered isoform whose ATPase activity and ability to translate actin filaments in a motility assay (38) is thrice that of all other isoforms, were unchanged.

The cause for the increased V_o induced by sensitization was increased phosphorylation of the regulatory MLC20 by the increased total activity of MLCK that we detected (39).

Alterations in the mechanical motor. While motor MLC20 phosphorylation could account for increased contractility, this may be only a part of the story; another part of the explanation might involve a change in the molecular motor itself. In a study by Stephens and Jiang (40), myosin heavy chain molecules were separated from both control and sensitized ASM. Actin filaments were obtained from chicken gizzard and employed to carry out a conventional motility assay (5, 7, 40, 41). Purified myosin molecules were made to adhere to a nitrocellulose-labeled cover slip using an interposed antibody. The MLC20 in the myosin head was fully phosphorylated using ATPS, or by predetermined amounts of the catalytic subunit of MLCK. Actin filaments labeled with tetramethylrhodamine-phalloidin were perfused onto the cover slip. As no ATP was present in the solution, rigor crossbridges were formed. When ATP was added the crossbridges cycled, movement of the actin filaments could be monitored, and the velocity of translation of the filaments measured. From the measurements the maximum velocity of translation (essentially zero-loaded velocity) was obtained (see Figures 11 and 12).

View larger version (133K): [in this window] [in a new window]   Figure 11.   Images of actin filaments in motion. ATP has been added and actin filament translation is occurring. Using computer software techniques, the maximum velocity of filament translation can be computed. Reprinted by permission from Reference 40.

View larger version (125K): [in this window] [in a new window]   Figure 12.   Similar images of actin filaments as shown in Figure 11. However, further time has elapsed and translation developed. From these frames the maximum velocity of translation is computed. Reprinted by permission from Reference 40.

No significant difference was found between sensitized and control myosins, from which we concluded that the molecular motor itself was unchanged and the increased velocity of shortening of sensitized myosin was due to its increased activation by increased total MLCK activity. The dose-response relationship between MLCK content and V_o was found to be sigmoidal and confirmed that motility of the actin filament was dependent on MLCK activity.

One of the unresolved questions in asthma relates to whether there is any primary change in contractility of the human bronchial smooth muscle. Current thinking is that any change is likely secondary to inflammation, or that while the muscle's contractility is itself normal, it shortens to a greater extent due to the increased amounts of agonists liberated by immune reaction. In vivo studies on human patients do not allow direct insight into cellular and molecular mechanisms because of the simultaneous operation of multiple variables. Studies conducted on postmortem or surgical tissue obtained from subjects with asthma are for the most part anecdotal (42), so no statistically significant conclusions can be drawn.

We have recently obtained preliminary results addressing the question. Professor Michel Laviolette of the University of Laval, Quebec, has developed an endobronchial biopsy technique by which he obtains satisfactory specimens that arrive in our laboratory within 18-24 h. Ninety-five percent of the muscle cells obtained enzymatically are viable as judged by trypan blue exclusion and by response to electrical stimulation. Similar specimens are obtained from healthy siblings and from human surgical specimens obtained from healthy parts of lobes resected from lung carcinoma.

Figure 13 shows results obtained from five preliminary experiments. The three bars on the left show statistically significant differences in maximum shortening ability between single BSM cells from asthmatic and normal subjects. The middle bar shows that, contrary to expectations, BSMCs obtained from surgical material from supposedly normal bronchi are in fact not normal (p < 0.05). Hence such material should not be used as a control.

View larger version (49K): [in this window] [in a new window]   Figure 13.   Mean shortening capacities of single bronchial cells obtained from human subjects and dogs.

The right panel demonstrates that single BSM cells obtained from dogs sensitized to ragweed pollen and their litter-mate controls show a statistically significant difference in shortening. We have also conducted studies of shortening from sensitized strips of BSM (24) and found statistically significant increases in shortening.

In the experiments on single cells from humans and from the dogs just described, V_o was also measured; it showed the same changes as L_max.

Airway smooth muscle in culture. We have recently observed that in cultured cells approaching confluence, when the 10% fetal bovine serum normally present in the nutrient medium is withdrawn, i.e., the cells are "starved," 30% of the cells show a new phenotype. Figure 14 shows cells at 75% confluence on Day 7 of culture (left panel). The right panel shows cells that have been starved for 10 d. Long cells are seen. These show the refractile edges typical of viable cells, and they demonstrate the ability to exclude trypan blue. Quantitative studies show that 30% of the cells in arrest convert to this phenotype, and furthermore they become "supra-contractile." This is, as far as we are aware, the first report of cells in culture regaining the ability to shorten considerably (36) and reversibly. These cells grow to the length of Type IIB cells, which is the same as that of cells in situ in mature ASM tissue. They respond well to electrical stimulation, as well as to stimulation by acetylcholine, histamine, and KCl. Shortening ability and V_o is almost double that of freshly isolated cells. This demonstrates the "arrested" cell has changed its biophysical phenotype and differs radically from the confluent cell.

View larger version (108K): [in this window] [in a new window]   Figure 14.   Morphology of cultured tracheal smooth muscle cells. The left panel shows airway smooth muscle cells in culture at 70% confluence. The right panel shows cells that have been "arrested" for 15 d, i.e., deprived of fetal bovine serum. Bundles of elongated cells are visible. Reprinted by permission from Reference 36.

Concomitantly there are changes in biochemical phenotype; for example, changes in sm-actin content are seen. Figure 15 depicts the result of a pilot study. On the seventh day in culture (CD7) the content is considerably reduced compared to content on day zero (CD0). On the 15th day of arrest (AD15) actin levels have increased to double those of CD0. Smooth muscle myosin heavy chain (SM1 and SM2) contents drop to almost undetectable levels at CD7 but recover to normal levels at AD15. These findings indicate that the increase in V_o cannot be accounted for by any increase in myosin heavy chain content since none occurred. As mentioned before, recently SM-B, a third myosin heavy chain isozyme, has been isolated, which has three times the ATPase activity of the other isoforms. Since suitable antibodies are not available but the mRNA sequence structure corresponding to the important 7-amino-acid (21 base pairs) structure in the amino-terminal head is known, we were able to synthesize the requisite primers to estimate the abundance of SM-B mRNA using RT-PCR techniques. Figure 16 shows bands representing SM-A (MHC1 and MHC2) and SM-B in freshly (F) isolated cells. From the molecular weight markers the difference between the two bands turned out to be 21 base pairs. On the very first day of arrest (Day 0) SM-B vanished. This was not unexpected as the cells were in a stage approaching confluence, at which time they were markedly de-differentiated and show almost no contractile activity. The surprise was the SM-B did not reappear even at Days 12 and 15 when the arrested cells were in their supercontractile state. Patently it is not change in distribution of isoforms, i.e., change in the endogenous nature of the ATPase, that is responsible for the increase contractility of the "arrested" cell.

View larger version (22K): [in this window] [in a new window]   Figure 15.   -Actin, MHC at CO, C7, and A15. The first three histograms show smooth muscle -actin content of freshly isolated (CO, i.e., culture at Day 0), cultured cells at Day 7 (in 70% fetal bovine serum) and cells that have been "arrested," i.e., serum-deprived cells, for 15 d. The next three histograms show similar data for smooth muscle myosin heavy chain.

View larger version (38K): [in this window] [in a new window]   Figure 16.   Smooth muscle myosin heavy chain (with seven amino acid insert)-mRNA data for airway smooth muscle cells. M represents the number of base pairs markers. F represents freshly isolated cells. The lower band represents smooth muscle myosin heavy chains I and II (SM-A)-mRNA bands, while the upper represents heavy chains with the seven amino acid insert (SM-B). At Day 0, the upper band vanishes and does not reappear even up until Day 15 of arrest. Reprinted by permission from Reference 36.

We next tested the hypothesis that increased contractility was due to a greater degree of activation of the ATPase by virtue of increased phosphorylation of its MCL20. Preliminary experiments showed that the content of sm-MLCK was reduced by 50% on the seventh day of culture (75% confluence in the presence of 10% fetal bovine serum). However, on the 15th day of arrest (15 d of incubation with 0% serum) MLCK content increased by 30-fold. Note that specific antibody against sm-MLCK was used and the molecular weight of the enzyme, as detected by Western blotting, was 140 kD. The 30-fold increase represented sm-MLCK alone. We do have antibodies against nonmuscle MLCK (nm-MLCK) and have seen that these label a quite different group of MLCKs whose molecular weights range from 204 to 214. Studies of MLC20 phosphorylation have been initiated, but no statistical statements can be made at this time. However, the considerable increase in sm-MLCK could result in sufficient increase in phosphorylation of MLC20 to account for the increased contractility of the arrested cell and support the hypothesis set out at the beginning of this paragraph.

Immunocytochemical Studies

The greatly increased content of MLCK led us to determine whether multiple isoforms existed and where in the cell they were located. Figure 17 shows a picture of freshly isolated (Day 0) cells treated with sm-MLCK antibody labeled with Cy-3. The MLCK is uniformly distributed throughout the cytoplasm, while the nucleus is free. The attenuated staining in the nuclear zone is perinuclear. Sodium dodecylsulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and Western blot analysis revealed the molecular weight of the sm-MLCK was 140 kD. Figure 18 shows a stained cell obtained from a 7-d culture in nutrient medium containing 10% fetal bovine serum. The nm-MLCK antibody kindly provided by Dr. Patricia Gallagher of the University of Indianapolis was used in conjunction with Cy-3. The nm-MLCK appears in linear cytoplasmic arrays co-localized with stress fibers. However, a high concentration of the enzyme was present in the nucleus as was proved by confocal microscopy. Furthermore, preliminary studies of isolated nuclei, using Samuel and colleagues' technique (43), demonstrate the presence of nuclear nm-MLCK. We have no idea as to the function of the enzyme in the nucleus. In karyokinesis, translocation of chromosomes in metaphase to the cellular poles is effected by a kinesin-microtubular system. One possibility is that the kinase is required for phosphorylation of a light chain in the kinesin molecule. For cytokinesis the enzyme would have to be outside the nucleus. It is worth noting that Guerriero and coworkers (44) documented the presence of MLCK in the fibrillar component of the nucleolus. Others have also demonstrated the nuclear location of the enzyme (32).

View larger version (74K): [in this window] [in a new window]   Figure 17.   Immunocytochemistry of freshly isolated (Day 0) airway smooth muscle cells. Antibodies against smooth muscle type myosin light chain kinase (sm-MLCK) were used. Cy-3 was used as the fluorescent probe.

View larger version (70K): [in this window] [in a new window]   Figure 18.   Immunocytochemistry of cells at Day 7 of culture. Antibodies against nonmuscle MLCK (nm-MLCK) were used; Cy-3 was the fluorescent probe employed.

	CONCLUSIONS

TOP ABSTRACT INTRODUCTION DEFINITION OF CONTRACTILITY MECHANICS OF THE ASTHMATIC'S THE MECHANICAL PROPERTIES OF PHENOTYPIC HETEROGENEITY OF CONCLUSIONS REFERENCES

The presence of normally cycling crossbridges, which are activated early in contraction, and latchbridges, which are activated later in contraction in airway smooth muscle, conforms to what is seen in most smooth muscle cells. An interesting consequence of the time-dependence of activation of these bridges is that almost 90% of muscle shortening is complete in the first 3 s of a total duration of contraction of 10 s. This highlights the importance of studying the early phase of shortening. One application of this in asthma research has been the finding that the increased extent of shortening of sensitized airway smooth muscle occurs entirely within the first 3 s of contraction. Another implication is that V_o becomes an important determinant of maximum shortening capacity. Since 90% of the total shortening has to be completed in 3 s, the muscle must be able to shorten rapidly enough to achieve this.

Not only is the load an important determinant of shortening but so is the nature and history of loading. For equivalent degrees of loading, auxotonic shortening is less than isotonic.

An important discovery has been the role of plasticity in smooth muscle contraction. Ford's group has shown that when length-tension relationships are measured in a repeatedly stimulated tracheal smooth muscle segment of fixed length, maximum active tension is independent of length over a wide range extending from 0.5 l_o to 1.5 l_o. They ascribe this behavior to plasticity of thick filament development, with new filaments being laid down in series during the course of contraction.

With respect to asthma, the still outstanding question is whether bronchial smooth muscle contractility is altered, or if the muscle is made to shorten more, simply by increased amounts of agonists liberated from the mast cells and other cells by the immune reaction. The major problem has been the great difficulty in obtaining bronchial smooth muscle tissue in adequate amounts from sufficient numbers of asthmatic subjects and healthy siblings. We, as well as Laviolette, have succeeded in achieving this, and the evidence is fairly strong that contractility is indeed significantly increased in bronchial smooth muscle from subjects with asthma. This discovery was made possible by developing techniques to measure single cell contractility.

A development with far-reaching consequences has been the discovery that considerable phenotypic heterogeneity exists at bronchial and tracheal smooth muscle cell level. This dictates that studies designed to elucidate pathogenesis of disease, or physiologic regulation at molecular level, must be conducted on single cells because the assumption that all the cells in a strip of muscle are homogeneous is no longer tenable. We have reported the presence of three types of cells in bronchial smooth muscle: Type I, which are short (100 µm length) and shorten by 25%; Type IIA, which are long (200 µm) but shorten by 25%; and Type IIB, which are 200 µm long and shorten by 65%. Undoubtedly, biochemical heterogeneity must also exist and remains to be investigated.

Finally, the discovery of a new supercontractile phenotype in culture is reported. When, in the course of airway smooth muscle cell culture, the cells are starved by culturing them in nutrient media from which fetal bovine serum has been removed, by the third to the 50th day of arrest long fusiform cells appear in bundles. On Day 15 of arrest, 25% of the cells are transformed. They are fully viable and show an unloaded shortening capacity and velocity that is double that of freshly isolated (Day 0) cells; in effect they become supercontractile. Biochemical studies show that the enhanced contractility is not due to any change in the intrinsic enzymatic properties of the myosin heavy chain isozymes but an increased degree of stimulation of the heavy chain enzyme by increased phosphorylation of its 20-kD MLC brought about by a 30-fold increase in MLCK content.