Contraction of smooth muscle in visceral organs is modified by structures external to the muscle. Within muscle tissue itself, connective tissue plays an important role in force transference among the contractile cells. Connections arranged radially can affect contractile mechanics by limiting tissue expansion at short lengths. Previous work suggests that increased stiffness at extreme shortening is due to such radial constraints. Two approaches to further study of these effects are reported. To increase radial constraints, very thin Silastic bands were placed loosely about strips of canine trachealis muscle at rest length. The strips were allowed to shorten under light afterloads, expanding until restrained by the bands. Subsequent removal of the bands allowed increased shortening, with less increase in stiffness at short lengths. Related isometric effects were observed. To reduce constraints, muscle strips were partially digested with collagenase. Compared with control conditions, this treatment permitted further shortening, with less increase in stiffness at short lengths. These results emphasize the role of extracellular structures in determining mechanical function of smooth muscle.
- smooth muscle contraction
- tracheal muscle
- connective tissue
in a large and complex organ such as the lung, smooth muscle must operate within a complex mechanical environment in which cartilaginous and parenchymal attachments modify its contractile functions. Many studies have examined the effect of forces and influences arising from various extramuscular components on the mechanical function of pulmonary smooth muscle (6, 7, 18, 21, 36). However, the mechanical performance of smooth muscle tissue itself, even when freed from the mechanical constraints of surrounding tissue, involves complex interactions between the contractile cells and the connective tissue of its own extracellular matrix. Measurements of cellular mechanics in a multicellular preparation, or the ability to predict the behavior of a multicellular preparation from the properties of its constituent cells, are limited by such interactions. In those smooth muscles that undergo considerable shortening as part of their normal physiological function (e.g., bronchial muscle, urinary bladder, or uterus), the effects of tissue length on the contractile properties of the whole tissue or organ are an important factor in determining the overall function.
Many mechanisms have been proposed to explain the length dependence of smooth muscle mechanical behavior, including cellular processes such as length-dependent activation (i.e., calcium entry and myosin phosphorylation) (12, 14, 15, 22), length-dependent rearrangement of the cellular constituents of the contractile mechanism itself (5, 13,37, 43), and the presence of length-dependent internal mechanical loading (16). Much evidence supports these mechanisms, all of which are based on processes that are primarily intracellular. The present paper addresses an additional mechanism that is related to extracellular structures and has not yet been studied extensively. In this regard, Gabella and Raeymaekers (11) describe the modified shortening behavior of collagenase-treated taenia coli smooth muscle, and Bramley et al. (2) have shown that collagenase treatment of isolated strips of bronchial muscle makes them capable of additional shortening. This finding differs from that of Ma and Stephens (20), who reported a decrease in force development with no increase in shortening. However, these results all imply that an important component of the mechanical environment of smooth muscle cells is provided by their extracellular matrix and cell-to-cell interactions.
The experiments in this paper address the question of extracellular mechanical constraints and are guided by a set of assumptions and proposed interactions that can be termed the “radial-constraint hypothesis” (26, 27, 32). They are designed both to provide further tests of the validity of the hypothesis and to explain features of the behavior of isolated smooth muscle in general. The hypothesis holds that a strip of smooth muscle, when very lightly loaded, will contract isotonically at approximately constant tissue volume. As the extreme of shortening is approached, the tissue must expand significantly in a radial direction to accommodate constant volume. This expansion is counteracted by connective tissue acting in a radial direction. The strain induced in this radial tissue acts as a load on the contractile apparatus and causes shortening to be limited, and it also causes the axial stiffness of the preparation to rise as cross bridges are recruited to bear the additional internal load. Thus the hypothesis defines a way in which the transverse (or radial) properties of the tissue can be investigated by making measurements of shortening and stiffness that are confined to the axial direction of the tissue strip.
Two approaches are reported here. In the one case, additional radial constraints were added (reversibly) to a contracting muscle strip, and the effects on the contraction were characterized. In the second case, radial constraints were removed (irreversibly) by mild digestion of the intact tissue with collagenase, and a similar mechanical characterization was done. In both cases, the results were consistent with the general predictions of the hypothesis.
All of the experiments described here were performed on isolated canine trachealis strips. Mongrel dogs were anesthetized with pentobarbital sodium and were rapidly exsanguinated. A 10- to 15-cm-long segment of extrathoracic trachea was quickly removed and placed in a physiological saline solution of the following composition (in mM): 125 NaCl, 4.7 KCl, 2.5 CaCl2, 1.2 MgSO4, 15.5 NaHCO3, 1.2 KH2PO4, and 11.5 glucose. The solution was bubbled throughout the experiment with 95% O2-5% CO2 to maintain a physiological pH. The cartilaginous rings were cut at both sides, and the trachea was pinned out in a dissecting dish. The muscle area was cleaned of epithelial and adventitial tissue. Under a dissecting microscope, small strips of muscle tissue (∼0.75 mm diameter and 8–12 mm long) were cut from the muscle sheet, with care being taken to follow the natural division of the tissue into discrete fiber bundles. To ensure a low-compliance (and removable) attachment to the experimental apparatus, the ends of the strip were clamped in aluminum foil cylinders as previously described (24). Direct measurements of total system compliance, which included that of the force transducer and all other mechanical components, amounted to 0.93 μm/mN. This was equivalent to <0.5% of the muscle length at maximal force (e.g., 50 mN) at optimum length (L o) (e.g., 10 mm).
After the tissue was mounted to the extension arms of the apparatus, it was extended by adjusting the micromanipulator that bore the force transducer until a small force (∼1–2% of the anticipated maximum) was recorded. This length was designated rest length (L r) and was ∼10% less thanL o. This procedure was used because the experimental protocols required that passive force be kept to a minimum. After the tissue was mounted, the muscle bath, mounted on a rack-and-pinion assembly, was elevated to immerse the muscle in circulating, temperature-controlled, and oxygenated physiological saline solution. Muscles were stimulated by using platinum electrodes along either side of the tissue, with supramaximal voltage pulses of alternating polarity at a frequency and voltage previously determined to produce a maximal mechanical response.
The protocols described below required close control of muscle force and/or length, along with continuous measurement of the muscle stiffness under both isotonic and isometric conditions. These requirements were met by using a force-clamp system built around a Cambridge 300H Ergometer with external force feedback. The muscle was attached at one end to the lever of the driver motor, while the other end was attached to the active element of a photoelectric force transducer with a natural frequency of 670 Hz and a linear compliance of 0.28 μm/mN. A dedicated digital computer system [suggested by the software design of Smith and Barsotti (40)] was used to provide feedback for isotonic force control. The output of the force transducer was compared with an adjustable reference voltage to produce isotonic conditions. Changes in experimental conditions were controlled by signals from a second digital computer system that also digitized and stored the experimental data.
Dynamic muscle stiffness was measured, as in previous studies on other tissues (23, 27, 31), by applying a sinusoidal length oscillation (32–100 Hz, amplitude <0.5% L r) to the muscle via the driver motor and analyzing the resulting force perturbations, with stiffness defined as the ratio of the force change to the length change (dF/dL). This system allowed continuous measurement of muscle stiffness, a critical parameter in these studies, under both isotonic and isometric conditions. At the perturbation amplitudes used, there was no discernible effect of the oscillations on the muscle performance (24).
Before control and experimental data were obtained, a preparation was equilibrated with a series of electrically stimulated isometric contractions (usually 5–10, at 5-min intervals) until a constant peak isometric force was obtained. During the subsequent control period, isometric (atL r) and isotonic contractions were alternated. An isotonic contraction (see Figs. 2 and 6, for example) usually began with a brief period of isometric activity. As force began to rise, the force-clamp system was switched to the isotonic mode, with the afterload controlled at a level very close to zero force (<5% of peak isometric force). The stimulus duration was set to allow complete shortening before the stimulus was turned off.
To produce an artificial radial constraint of shortening tissue, five to six very thin (<0.2 mm) restraint bands cut from Silastic tubing were placed around the muscle strip before applying the aluminum foil attachment fixtures (see Fig. 1.) Care was taken to ensure that the bands fit loosely around the strip at theL r and that they were evenly spaced (there was very little lateral migration of the bands during the experiment). After the control contractions on the banded strip were finished, the muscle was removed from the apparatus, keeping the foil attachment fixtures intact. The bands were carefully dissected away, and the strip was remounted to the apparatus by using the same mounting holes as in the control situation. After the remounting, a set of measurements complementary to those done in the banded condition were carried out. The physical manipulations involved in these procedures did not appear to have a detrimental effect on the performance or viability of the preparations, although small changes in the shortening velocity and the isotonic stiffness at the onset of shortening were noted (see results). Twelve muscles, with a meanL r of 9.75 ± 0.62 (SD) mm and weight of 4.8 ± 1.6 mg, were studied in this manner.
To reduce the (hypothetical) degree of radial constraint, 17 muscles (mean L r, 9.57 ± 0.54 mm; mean weight, 5.9 ± 1.5 mg) were subjected to mild collagenase digestion, by using collagenase type III (Worthington Biochemical, Freehold, NJ) at a mean concentration of 539 ± 9.1 U/ml. The muscle was left attached to the experimental apparatus, and an auxiliary bath (volume 3.0 ml) with thermal connections to the experimental tissue bath was used to maintain the muscle at ∼37°C during the digestion. During the digestion period, which averaged 21.4 ± 0.44 min, the solution was gently stirred, but care was taken to prevent any mechanical perturbation of the muscle preparation. At the end of the digestion period, the auxiliary bath was removed and was replaced with the main experimental muscle bath. Washing of the preparation was accomplished by the continuous circulation of physiological saline in the larger volume (35 ml) of the muscle bath. Sham digestions, using the above protocol but without collagenase, were performed as controls. There was no significant change in muscle properties due to the experimental manipulations.
All experimental protocols were arranged so that each muscle served as its own control. This procedure minimized animal-to-animal variation and allowed the detection of small but significant variations with the use of a paired t-test procedure (SigmaStat; Jandel Scientific).
Isotonic behavior of constrained muscle.
Figures 2 and 3show the isotonic contraction of the same tracheal muscle strip with and without the presence of applied constraints (5 bands, 1.2-mm inside diameter, 0.2-mm thick, evenly spaced). Compared with the control condition (Fig. 2 B), the constrained tissue (Fig. 2 A) shortened less (by ∼15% of L r), and the shortening was accompanied by almost twice the increase in peak isotonic stiffness. During the isometric and early isotonic portions of the contraction, there was no apparent effect of the constraint bands on muscle behavior, although there were some changes in the initial isotonic stiffness and shortening velocity. For the data set summarized in Fig. 4 and Table1, the mean stiffness at the onset of isotonic shortening was 0.061 ± 0.033 mN/μm with the bands applied. This fell to 0.051 ± 0.020 mN/μm on their removal. Similarly, initial shortening velocity for the banded muscles was 2.650 ± 0.746 mm/s, whereas that for the same muscles unbanded was 2.370 ± 0.706 mm/s. This represented a 9.0% decline in velocity. A paired t-test showed that this decline was statistically significant (P = 0.006). However, since the paired before and after measurements spanned a significant amount of time (up to 2 h), this slight fall could be attributed to the expected rundown of the preparation rather than to a consistent effect of the constraint bands. In those instances in which the entire curves were determined, the force-velocity relationship (not shown) was unchanged by the removal of the constraint bands, because all velocities were measured at lengths close toL r, where the constraint bands were still slack. The shortening length at which the effect of the bands did become evident depended significantly (for a constant size of band) on the size (mass) of the muscle (P < 0.001).
A closer view of the effect of applied constraints is shown in Fig. 3, in which the muscle stiffness is plotted as a continuous function of length during isotonic shortening. Early in the contraction the curves do not differ greatly, but after a few millimeters of shortening, the stiffness of the banded muscle (□) begins to increase rapidly and reaches its highest value at the end of the (somewhat restricted) shortening. The same preparation, with the bands now removed (○), is free to shorten further, and less stiffness is developed as shortening continues. These results are summarized in Fig. 4 and Table 1, in which 47 paired contractions of 12 muscle preparations are compared. The differences between both parameters, as determined with a pairedt-test, are highly significant (P < 0.0001). Expressed in other terms (and with the same degree of statistical significance), the banded muscles shortened to 44.1% ofL r, whereas the same muscles, after unbanding, shortened further to 31.5% ofL r. Concomitantly, the banded muscles showed a stiffness increase during shortening of 32.2%, whereas the same muscles unbanded showed an increase of 17.8%. The relatively small changes in shortening velocity and initial stiffness, as mentioned earlier, may reflect temporal changes in the muscle unrelated to the presence of the constraint bands. Note that 12 pairs of values are shown for the individual muscles in Fig. 4. For clarity in plotting, multiple determinations on individual muscles were averaged to provide one set of data per muscle. This provisional data pooling did not result in any significant changes in the statistical parameters. The statistical calculations in Table 1and the box plots in Fig. 4 were based on all 47 individual comparison pairs. The unbanded-to-banded ratio [Ratio (U/B)] in Table1 is arranged to show the effects of removing the bands; it is thus analogous to the results of enzymatic digestion to be reported below.
Isometric behavior of constrained muscle.
The relationship between the length and diameter of a muscle (with constant volume) should be relatively independent of whether the muscle is contracting under isotonic or isometric conditions. A test of this assumption is shown in Fig. 5. Conventional isometric length-tension curves were constructed for a muscle strip before and after removal of the constraint bands (Fig.5 A). The data points have been fitted with arbitrary second-degree polynomial functions, and the 95% confidence intervals are shown. It is apparent that the presence of the constraint bands reduced the muscle force at the shorter lengths while having no detectable effect at lengths nearer toL r. Lightly loaded isotonic contractions were also made under the same conditions, and a plot of their length vs. stiffness relationship is shown in Fig.5 B. Although it was difficult to define precisely the length at which the two length-tension curves diverged, this parameter could be estimated by choosing the point at which the 95% confidence intervals no longer overlapped. When this point was compared with the divergence shown by the length-stiffness plots in Fig. 5 B, a close correspondence was found. For five muscles of differing weights that underwent this isotonic-isometric comparison protocol, the lengths at which the isometric length-force and isotonic length-stiffness curves diverged showed a significant correlation (Fig.5 C).
Effects of partial digestion on shortening and stiffness.
Figure 6 shows the results of a 17-min digestion (with 270 U/ml of collagenase type III) on the isotonic behavior of a muscle strip. Compared with the control record (Fig.6 A), the digested strip shortened further but became considerably less stiff at the peak of shortening (Fig. 6 B). Note also that the stiffness at the beginning of isotonic shortening was greater for the digested muscle. The initial isotonic shortening velocity was also increased by a small, but statistically significant, amount.
The contraction data from Fig. 6 are further illustrated in Fig.7 (compare with Fig. 3 for banded muscle). For use in analyzing multiple contractions, data values for peak shortening (1), peak isotonic stiffness (2), initial isotonic stiffness (3), and the initial shortening velocity (not shown in Fig. 7) were measured on paired contractions (intact vs. digested) of 17 muscle preparations. Because of the potentially destructive effects of the enzyme treatment on muscle viability, comparisons were made only between the last predigestion contraction and the first postdigestion contraction. Thus there was only one contribution to the data set from each muscle (in contrast to the case for the banded muscle experiments). Data were analyzed for significant differences with a pairedt-test (see Table 1 and Fig.8). Although the effect of digestion on peak shortening was not very large, it was highly significant (P = 0.0003). The effect on peak isotonic stiffness was considerably greater and was also highly significant (P = 0.0001), as was the effect on the initial isotonic stiffness, although there was greater variation in the initial stiffness data. Postdigestion shortening velocities (at the beginning of the isotonic phase of the contraction) were consistently and significantly higher that the predigestion control values, an effect that may reflect a reduced internal load on the contractile cells (seediscussion). Statistical analysis of the stiffness during the rise of isometric force from the data set reported in Fig. 8 (before the sudden step to the afterload; see Fig.6, bottom set of traces) showed that it was not significantly affected by the digestion (pairedt-test).
Longer term effects of mild digestion.
Figure 9 shows data from a muscle strip digested for 22 min with 552 U/ml of collagenase type III. Compared with the control record (left), the peak shortening and the initial velocity of shortening were maintained at a nearly constant level for >30 min after the washout of the enzyme preparation, although some of the later contractions did not show the enhanced shortening. However, the peak stiffness associated with the maximal shortening showed a steady and marked decline over the same period of time.
The decline in isometric muscle function was also followed after the period of digestion (unpublished observations). In a typical strip, in which contractions were made every 6 min, the peak isometric force fell steadily in an approximately exponential manner. Increasing the interval between contractions decreased the rate of decline, but when the data were analyzed on the basis of each contraction, the rate of decline became constant. This indicates that the decline was due primarily to mechanical damage to weakened connective tissue caused by the force generated by the cells. Peak isometric force and stiffness fell proportionately (i.e., the ratio of stiffness to force was preserved, as noted in the previous section), indicating a common locus for the elements bearing the developed force and those providing the axial resistance to elongation.
Summary of major findings.
Mechanically constraining the increase in diameter of a shortening muscle resulted in a decrease in the amount of shortening, coupled with an increase in the stiffness at the shorter lengths. Partial digestion of a muscle strip produced the opposite effects, suggesting that the enzyme treatment has reduced internal constraints on tissue expansion. Smaller but significant effects were found on the initial isotonic shortening velocity and on the initial isotonic stiffness. These results will be discussed in the context of the radial-constraint hypothesis, with its inherent implications for the function of smooth muscle in organs where muscle is subject to large amounts of shortening.
The role of muscle in any organ whose function requires an active dimensional change is obviously critical. As is the case in many other organs, the muscle components of the lung shorten on their own, but they also are stretched and released by external forces during the respiratory cycle (7). Even when airway muscle is studied in isolation, the contractile activity of its cells is subject to significant modification by connections and interactions within the muscle tissue itself. It might, therefore, appear that study of isolated muscle cells would be a fruitful and relatively uncomplicated source of basic information about the mechanical function of smooth muscles. In fact, this approach has yielded much of value in our understanding of muscle function and continues to do so, despite the considerable technical difficulties involved. However, as a number of structural studies have emphasized, the mechanical conditions facing the contractile apparatus of isolated cells are significantly different from those experienced in an intact tissue (8-10, 19). In particular, a cell within a tissue is both a generator and a transmitter of mechanical activity, and mechanical connections are made at numerous locations on the lateral surfaces of the cell via membrane-associated dense bodies and their intra- and extracellular connections (1, 4, 42) rather than just at the ends of the cell. A cell in isolation can be connected by its ends to experimental apparatus, but such a connection presents an arrangement much different mechanically from the natural tissue-based connections. For these reasons, it would appear that there are significant advantages to developing ways of making detailed mechanical studies of smooth muscle tissues in a relatively intact state, with the goal of drawing inferences about cellular mechanical function. The validity of such inferences must rest on an understanding of both tissue structure and the mechanical relationships that the structural arrangements would determine. The present paper is an attempt to further such an experimental and analytical framework.
Assumptions of the radial-constraint hypothesis.
Important elements of this hypothesis have been addressed in experiments on other tissues (26, 27, 29, 30), and a brief synopsis of its salient features has been given in the introduction to this paper. Because the essence of the proposal is that the length-dependent increase in stiffness is due to geometrical, not biochemical, factors, it should be applicable to a wide range of smooth muscle tissues. One element of the hypothesis, however, does concern activities occurring within the cells. Although there have been recent cautions expressed in the case of skeletal muscle, it is generally accepted that the stiffness of contracting muscle is related to the size of the active cross-bridge population. The close relationship between developed force and stiffness in isometrically contracting smooth muscle has been demonstrated for a number of smooth muscles (13, 17, 23, 34, 44), and deviations from this relationship have been used to provide evidence for particular conditions of the contractile system (13, 25, 31). In the present case, because the force-clamp system keeps the external force low and constant, it is assumed that the measured increases in axial stiffness represent increased cellular cross-bridge activity, which would vary in response to a mechanical load developed within the tissue. An attempt has been made to calculate the size of such an internal load, for instance in ovarian ligament muscle (30), on the basis of the measured relationship between force and stiffness under isometric conditions. At short lengths and low afterloads, it was shown to be an appreciable fraction of the externally manifested force. These arguments assume that the size of the active cross-bridge population in smooth muscle is capable of rapid adjustment in response to a changing load, whatever its source. Unpublished experiments from this laboratory show that this is the case for trachealis muscle; a change in applied load causes a rapid adjustment to the force-velocity-length conditions appropriate to the new load. These properties also hold for step increases in load. The rapid adjustment in force-dependent muscle stiffness that occurs when external loading conditions change (for example, in the transition from isometric to isotonic contraction; see Fig. 2) further suggests that rapid changes in the size of the cross-bridge population are possible and may play a role in the adaptation to changing internal loads at constant external force.
The internal load revealed by the radial-constraint measurements is probably not to be identified with the intracellular internal loads revealed by measurements of single cells (16), because, in that case, the load was proportional to cell length over a wide range and was not associated with a significant rise in stiffness. The radial-constraint internal load is highly nonlinear with muscle length (as predicted by the hypothesis), and thus it is likely to be tissue based. The presence of other (cellular) internal loads in tracheal muscle (see Ref. 39) is not ruled out by these experiments. In fact, the continuous fall in velocity during isotonic shortening even at moderate lengths (Figs. 2and 6) suggests that intracellular loads do influence the speed of muscle shortening.
Results from banded muscle strips.
The effects of externally applied constraints in limiting the expansion of shortening muscle strips are, to a first approximation, qualitatively consistent with the hypothesis under investigation, but some caution must be applied to interpretation of the results. The constraint bands do not contain the tissue uniformly, as would be the case if the muscle were contained in a continuous rigid cylinder. To allow longitudinal freedom of movement, the bands were made very thin (<0.2 mm), thus minimizing mechanical interaction between adjacent bands. However, unconstrained tissue would be allowed to expand between adjacent bands, and the expected effects would be reduced. In addition, for smaller muscles, their own internal radial connective tissue elements would begin to come into play before the constraint bands produced much of an effect. Despite these quantitative shortcomings of the constraint-band experiments, the results do support the basic qualitative predictions of the hypothesis.
The agreement between the isotonic and isometric effects of constraining bands (Fig. 5) makes it apparent that, for full isometric force to be developed, the muscle strip must be free to assume the appropriate diameter when active. These results imply strongly that one of the determinants of the isometric length-tension relationship in intact smooth muscle is lateral connections within the tissue that prevent full external (axial) expression of internally generated force. This is further supported by measurements which show that isometric muscle stiffness undergoes less of a relative decline in force at short lengths than does the isometric force (23). Both of these observations emphasize that, in addition to cellular mechanisms such as myofilament overlap and varying activation, the overall tissue geometry must be considered as a significant contributor to the isometric length-tension relationship.
Mechanics of partially digested tissues.
This portion of the experimental protocol was designed to address the question of radial constraints by attempting to remove them, rather than by adding them as in the first set of experiments. Efforts were made to keep the protocols as similar as possible, although the effects of the digestion made it impossible to complete a reliable length-tension curve on the digested muscle. In several instances, both the banding and digestion protocols were carried out on the same muscle strip. This practice was not generally used, however, because of the possibility of the deterioration of the preparation, especially after the period of digestion.
The continued decline in isometric force after digestion (seeresults) and the decline in peak isotonic stiffness (Fig. 9) might be viewed as such a decline in viability. However, other measures of contractility, such as the peak velocity of shortening or the maximal degree of shortening, do not show a steady and continued decline. These findings may have two explanations. If the activity of the collagenase had continued even during the period of washout after digestion, then the decline could represent further proteolysis of supporting structures. However, it is also likely that continued mechanical activity, which would produce internal longitudinal (isometric conditions) or radial (isotonic conditions) mechanical stresses, resulted in physical disruption of internal connections weakened by the digestion. This is borne out by the observed exponential decline in peak isometric force (seeresults), which suggests that the amount of internal disruption experienced with each contraction is proportional to the force produced in the preceding one. Thus the decline becomes less as the self-generated disrupting force becomes less and the weaker connections have been progressively broken.
Viewing the digestion results in light of the findings from the artificial constraint experiments, it is apparent that different degrees of constraint were present in the two cases. Removal of the natural constraints by digestion did allow the muscle strips to shorten further. Although the amount of additional shortening was not large compared with that seen in the banding experiments, the effect was consistent, with a very high statistical significance. This smaller increase in shortening could be explained in two ways. In the first place, the natural radial constraints were likely to be less robust than those provided by the Silastic bands; secondly, the degree of proteolysis was probably not uniform across the diameter of the strip. With digestion conditions kept mild enough to preserve the mechanical integrity of the preparation, a considerable portion of the radially disposed connective tissue would have escaped degradation. The small increase in shortening velocity after digestion (see Fig. 8) may have been due to reduced internal load associated with loss of connective tissue, but it is also possible that the period of mechanical quiescence during the digestion period allowed a more complete recovery of contractility than did the usual 6 min between successive contractions. However, there was no significant correlation between the increased velocity and the degree of enhanced shortening in individual muscles.
The peak stiffness data also showed different degrees of sensitivity to the experimental protocols. Whereas the artificial constraint bands almost doubled (1.85 times) the peak stiffness at the shortest lengths, the stiffness of the digested muscle fell to about three-quarters (0.793 times) of its control value. These differences are again probably due to the factors discussed in regard to maximal shortening. It could also be argued that the digestion process, in addition to its effects on the radial connective tissue elements, also affected connective tissue elements in series with the muscle cells, and this might account for the reduced peak isotonic stiffness. However, because the ratio between force and stiffness was maintained during the degradation of the tissue and because the axial force was kept very low, this factor should have minimal effect. A quantitative analysis performed on a mathematical model of the radial-constraint hypothesis (29) revealed that the decrease in axial stiffness was considerably more sensitive to degradation of radial (circumferential) elastic elements than to longitudinal elements.
The observed increase in initial isotonic stiffness shown by the digested muscle appears anomalous at first. However, it may be explained by a net shift, caused by the cutting of critical internal connections, from a more series-connected to a more parallel-connected orientation of the connective tissue throughout the muscle strip. Such behavior has been demonstrated for an artificial network of elastic elements and their interconnections (3) and appears to be a phenomenon that is a general property of some types of networks.
Implications of the findings.
The present study has attempted to view tracheal smooth muscle as an integrated tissue, intact but removed from the mechanical constraints of its usual attachments in the trachea. The aim of the isolation was to permit a focus on mechanical constraints inherent within the structure of the tissue itself. The degree of shortening of tracheal muscle in situ is limited by its connections to the cartilaginous rings of the trachea, and it is possible that this muscle in vivo never experiences the extreme shortenings that are associated with the sharp increase in axial stiffness. However, it is possible that muscle at more distal regions of the respiratory tree does experience shortening sufficient to produce an increase in internal load (2, 21). In any event, the presence of the phenomenon in isolated muscle does reveal features of tissue assembly that are not mechanically evident under conditions of less extreme shortening but which nonetheless are features of its functional arrangement and may play a role in airway mechanics.
Most studies of the mechanical properties have been confined to the analysis of forces and connections arranged in the axial direction of the tissue. Reports of the measurement of the transverse elastic properties of isolated muscle tissue, especially smooth muscle, are not common in the literature. Tsuchiya and co-workers (41) have used an ultrasound technique to measure transverse stiffness in skeletal muscle, and there have been attempts to use ultrasound to make dimensional and mechanical measurements on vascular muscle in situ (38). However, the tissue structure of smooth muscle is quite unlike that of skeletal muscle, and in the case of the vascular muscle, geometric factors of the vessel wall obscure the properties of the muscle itself. This present study offers an indirect but mechanically straightforward means of investigating multidimensional mechanical properties of muscle tissue from a number of organs and with a variety of experimental goals. In addition to revealing the presence of radial forces in intact tissue, these experiments suggest the possibility of physically mapping the mechanical interconnections of the tissue by selective disruption of the connective tissue with a variety of histolytic agents (33). Work in other laboratories has also demonstrated the continued viability of smooth muscle tissues subjected to partial digestion, and enzymatic and structural methods have been used to investigate the effects of interactions between cells and tissue (2, 8, 10, 11). Subsequent challenging of the weakened tissue with specific mechanical maneuvers designed to produce selective internal stresses (33) has the potential of being able to define specific types of physical interconnections within the tissue.
It is possible that a proteolysis-based alteration of airway muscle, analogous to the present digestion experiments, may take place in the asthmatic human lung. The presence of the enzymes chymase and tryptase, which are released from mast cells during an inflammatory response (35), may contribute to degradation of components of the extracellular matrix in airway smooth muscle. This could have the effect of lowering internal loading in the affected muscles, giving them the ability to shorten more completely in response to a given stimulus. In other smooth muscle tissues that undergo relatively rapid and substantial remodeling (such as in the pregnant, parturient, and postpartum uterus), the techniques presented here offer the possibility of following in considerable detail the natural progression of changes in the tissue structure.
These studies have demonstrated the role that radial mechanical constraints, based in the connective tissue architecture, play in limiting the shortening of smooth muscle tissues. Length-dependent isometric function was also shown to be subject to similar constraints. Such internal mechanical factors may play a role in the physiology of the normal lung and may also be involved with the function of airway muscle in disease states. Enzymatic alteration of the integrity of the extracellular matrix connective tissue was compatible with continued viability and shows promise as a method for further studies of the relationship between tissue structure and mechanical function.
The author thanks Dr. Susan Gunst for generous provision of tracheal muscle tissue and for many helpful discussions.
Address for reprint requests: R. A. Meiss, Depts. of Physiology and Biophysics, and Obstetrics and Gynecology, Indiana Univ. School of Medicine, Rm. 111, Med. Research Facility, 1001 Walnut St., Indianapolis, IN 46202 (E-mail:).
The author also acknowledges the financial support of the Dept. of Obstetrics and Gynecology, Indiana University School of Medicine.
A portion of this work has appeared previously in abstract form (32).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.
- Copyright © 1999 the American Physiological Society