Skeletal muscles are diverse in their properties, with specific contractile characteristics being matched to particular functions. In this study, published values of contractile properties for >130 diverse skeletal muscles were analyzed to detect common elements that account for variability in shortening velocity and force production. Body mass was found to be a significant predictor of shortening velocity in terrestrial and flying animals, with smaller animals possessing faster muscles. Although previous studies of terrestrial mammals revealed similar trends, the current study indicates that this pattern is more universal than previously appreciated. In contrast, shortening velocity in muscles used for swimming and nonlocomotory functions is not significantly affected by body size. Although force production is more uniform than shortening velocity, a significant correlation with shortening velocity was detected in muscles used for locomotion, with faster muscles tending to produce more force. Overall, the contractile properties of skeletal muscles are conserved among phylogenic groups, but have been significantly influenced by other factors such as body size and mode of locomotion.
- skeletal muscle
- shortening velocity
- tetanic tension
skeletal muscles are diverse in their contractile properties, with significant differences existing among and within various animal species. As such, skeletal muscles are striking examples of biological structures adapted for a specific function. Yet skeletal muscles are also highly conserved in terms of the molecular mechanisms responsible for producing muscle contraction (91, 99). Over the last 10 years, a considerable amount of data has been collected on the contractile properties of skeletal muscles from a diverse group of animals. In this study, the contractile properties of >130 skeletal muscles were analyzed to determine what broad trends could be observed that might give insight into the principles of skeletal muscle design. These muscles represent several distinct phylogenic lines, have varied functional demands, and come from animals spanning more than eight orders of magnitude in body mass. Although there have been several excellent reviews summarizing skeletal muscle properties, most focused on terrestrial mammals and none subjected the data to any type of statistical analysis (22, 54, 91, 95). In the present study, maximum shortening velocity (Vmax) and maximum tetanic tension (Po) were analyzed with respect to body mass, taxonomic group, mode of locomotion, and compared with one another.
The wide range of shortening velocities is one of the most prominent features that distinguishes muscle fiber types, with values in the present study ranging from ∼0.5 muscle lengths (L)/s to nearly 40 L/s. Vmax is often used to define different fiber types and depends primarily on the type of myosin heavy chain (MHC) expressed by a fiber (91). Fibers with different shortening velocities are found within individual organisms as well as among different species, with the contractile properties being matched to the specific mechanical needs of a particular muscle (86, 87). Integral to these mechanical requirements, muscle power is known to be a function of a muscle's force-velocity profile, with maximum power generally being produced at a V/Vmax of ∼30% (34,54, 86, 91). Shortening velocity has also been shown to be a major determinant of locomotion energetics, providing an explanation of why small animals use relatively more energy during locomotion than large animals (61, 84). Given the important linkages between shortening velocity and muscle function, it is of interest to examine whether any systematic trends exist among the shortening velocities from different skeletal muscles. Hill (42) predicted that shortening velocity should scale as an allometric function of body mass, with small animals possessing faster, more powerful muscles. This prediction has since been tested, with several studies finding that Vmax scales with a mass exponent between −0.11 and −0.20 (61, 88, 90, 100). Such scaling effects have been related directly to the different MHCs present in various muscle fibers (82, 91). The activities of glycolytic and oxidative enzymes that fuel locomotory processes also exhibit significant scaling effects in a variety of animals (19,30, 36). To date, the analysis of shortening velocity has been restricted to terrestrial mammals and it is unclear whether the observed scaling effects are applicable to animals with different phylogenic histories and modes of locomotion. In the current analysis, I examined whether there were any identifiable trends in Vmax with respect to body size, taxonomic group, and mode of locomotion.
Unlike Vmax, Po is generally considered to be constant among fiber types (91). Nevertheless, there have been inconsistencies in the determination of Po(91) and several studies documented significant differences related to fiber type (18, 37, 38, 60). Po is the product of the force produced per cross bridge, the number of cross bridges per unit area, and the muscle's duty ratio (proportion of cross bridges producing force) (91, 94). The type II myosins found in skeletal muscles have a lower duty ratio than other motor proteins as an apparent adaptation for increased speed (44); however, a lower duty ratio also results in lowered force production (44, 85, 94). In this context, the higher duty ratio in smooth muscles is thought to be responsible for the slower, more forceful contractions than those measured in skeletal muscles (7). Rome et al. (85) found that a disproportionate increase in cross bridge detachment rate relative to attachment rate in the toadfish swimbladder muscle (a high-frequency muscle used for sound production) resulted in a significant reduction in the duty ratio and thus force production. This example illustrates that as Vmax increases, the myosin attachment rate must increase in proportion to the rate of detachment or force production will be impaired (85, 94). It is not currently clear whether any systematic differences in Po exist among various skeletal muscles.
One of the contributions of comparative physiology has been the detection of underlying principles that unify biology despite the many physiological specializations observed in different animals (92). The goal of this study was to examine broad patterns that transcend the diversity in function often noted in skeletal muscle biology. The scaling of Vmax with body mass reveals a pattern more general than has been previously appreciated, applying not only to terrestrial mammals but to other terrestrial and flying animals as well. These results suggest that body mass has had a common influence on the shortening velocity of many locomotory muscles, irrespective of phylogenic group. Although maximum force production is more uniform than shortening velocity, a significant correlation was detected between Po and Vmax, with faster muscles producing slightly higher forces. In addition, some invertebrate muscles with long sarcomeres (6–15 μm) produce significantly more force than muscles from other animals. Taken together, these findings provide clues to explain some of the principles that influence skeletal muscle design.
MATERIALS AND METHODS
Published values for Vmax (L/s) and Po(kN/m2) were taken from the available literature for 72 species representing several different taxonomic groups, including mammals, birds, reptiles, amphibians, mollusks, insects, and crustaceans (Table 1). When not reported, body masses were estimated from other sources. In some instances, Vmax was not reported directly, but could be estimated from the reported data. For example, muscle strain and contractile frequencies in some reports were used to estimate shortening velocities. When shortening velocities were not explicitly maximum but were from animals engaged in locomotory activities, I assumed that power was being maximized and that this maximum was reached at 30% of Vmax. These estimated values are indicated in Table 1.
The relationship between Vmax and body mass was examined by log transformation of the data and using least-squares linear regression. Several analyses were performed, including an analysis with all of the data, as well as separate analyses with the data segregated by functional group (terrestrial, flying, swimming, and nonlocomotory muscles). When significant relationships were found for the different functional groups, the data for both sets were combined and analyzed using an analysis of covariance (ANCOVA) model to determine whether the regressions were significantly different from one another. When neither the slope nor the elevation were found to differ, the data were pooled and fit with a single regression. Normality was verified by examining frequency histograms of the data and residuals were plotted as a function of mass to ensure that the model was appropriate for the data. In addition, residuals were plotted as a function of experimental temperature to examine whether this additional factor could account for any variability in the data. These residual analyses identified experimental temperature as a significant predictor of Vmax; therefore temperature was used as second independent variable in multiple regression analyses. The effects of temperature per se were not of primary interest in this study. However, the wide range of experimental temperatures employed in these studies represented a confounding variable that obscured the scaling effects of interest. Including experimental temperature as a second variable in the regression analysis separated temperature effects from the effects of body size, resulting in a more meaningful estimate of the mass exponents.
The relationships between Po and body mass and between Po and Vmax were analyzed by linear regression as described above. Residual analysis of the regression of Po on Vmax revealed temperature as a significant predictor of Po, so temperature effects were partitioned by using a multiple regression analysis. When no significant regressions were detected, Po values were examined as a function of taxonomic grouping using a Kruskal-Wallis rank test, because the data showed significant departures from normality and equality of variances required for ANOVA. An analog of the Bonferroni pairwise comparison (77) was used to identify groups with significantly different rankings for Po (experiment-wise α = 0.05). Statview 5.0.1 (SAS Institute) was used for all statistical analyses.
A significant (P < 0.0001,r 2 = 0.34) relationship was found between body mass and Vmax for all muscles grouped together: Vmax = 10.8 · mass−0.17 (Fig.1 A). Further analyses of these data segregated by functional group (terrestrial, flying, swimming, or nonlocomotory muscles) indicated that the allometry was dominated by the muscles of terrestrial and flying animals (Fig. 1, B-D). Muscles from terrestrial animals showed a significant (P < 0.0001, r 2 = 0.61) relationship with body mass: Vmax = 15.9 · mass−0.25. A group of data from bird flight muscles measured in vivo was fitted with a separate line, because ANCOVA demonstrated that these Vmax values were significantly higher than the other flight muscles (P< 0.0001; Fig. 1 C). The flight muscles with these data omitted showed a significant (P < 0.0001,r 2 = 0.62) relationship with body mass: Vmax = 8.3 · mass−0.20 (Fig.1 C). The bird flight muscles measured in vivo also showed a significant (P < 0.024; r 2= 0.60) scaling relationship with body mass: Vmax = 38.6 · mass−0.13, but with an elevation more than four times that of the other fliers. The slopes of these two lines were not significantly different from one another (P > 0.63). ANCOVA revealed that the scaling relationships for muscles used in terrestrial locomotion and in flight (with the indicated bird data omitted) were not significantly different from one another. Therefore, these data were grouped together and analyzed by least-squares regression. These data showed a significant (P < 0.0001, r 2 = 0.65) relationship with body mass: Vmax = 10.2 · mass−0.20(Fig. 1 D). The swimming and nonlocomotory muscles showed no significant scaling effects (Fig. 1, E and F, respectively). Analyses of the residuals from all of the data and from the terrestrial/flying muscles indicated significant correlations with experimental temperature (Fig. 2,A and B). Adjusting for the effects of experimental temperature with multiple regression analysis reduced the mass exponent for all of the data from −0.17 to −0.11: log Vmax = 0.306 − 0.11(log mass) + 0.03(temp °C) (P < 0.0001; r 2= 0.57). Similarly, after the effects of different temperatures in the terrestrial/flying data were accounted for, the mass exponent was reduced from −0.20 to −0.12: log Vmax = 0.19 − 0.12(log mass) + 0.03(temp °C) (P < 0.0001;r 2= 0.78).
There was no relationship between Po and Vmaxfor all of the data together (not shown). However, there was significant (P < 0.039; r 2= 0.07) correlation between Po and Vmax for locomotory muscles, with faster muscles tending to produce higher forces: Po = 122 Vmax 0.14(Fig. 3 A). A residual analysis indicated that experimental temperature was also a significant predictor of Po (P < 0.03;r 2 = 0.07; Fig. 2 C); therefore temperature was entered into the regression to account for the influence of experimental temperature. Including temperature in the model resulted in an increase in the Vmax exponent from 0.14 to 0.28: log Po = 2.28 + 0.28(log Vmax) − 0.01(temp °C) (P < 0.003;r 2 = 0.18). There was no significant relationship between body mass and Po (Fig. 3 B). However, Po was found to be significantly higher in the crustaceans, amphibians, and mollusks than in the other taxonomic groups (Fig. 3 C).
It has long been recognized that smaller animals tend to have faster muscles than larger animals (42, 74). Hill (42) proposed that animals of differing size possess relative differences in limb dimension as a result of geometric scaling, predicting that shortening velocity should scale to match the natural frequency of the limbs, which in turn should scale as mass−0.33. Others have argued that animals do not scale geometrically, but that limb dimensions scale according to the principles of elastic similarity and that Vmax scales as mass−0.125 (74). The results presented here support the hypothesis that Vmax in terrestrial and flying animals scales very close to mass−0.125. When experimental temperature is included in the regression model, Vmax in the terrestrial and flying animals scales as mass−0.12. Likewise, shortening velocity scales as mass−0.13 in the bird flight muscles measured in vivo. These results are also consistent with other studies that have found Vmax to scale as mass−0.13 (90) or as mass−0.11(100) in skeletal muscles measured under uniform experimental conditions. An assumption common to current explanations of this allometry is that the natural frequency of limb movement is derived from the physical constraints of scale and that shortening velocity has been selected to match the natural frequency of the limbs (42, 61, 74, 88, 90, 100). Fairly consistent with this assumption, stride frequency has been shown to scale as mass−0.15 in running mammals (41), wing beat frequency in birds scales as mass−0.33 (81), and cycle frequency for a number of invertebrates scales approximately as mass−0.20 (34). Nevertheless, the assumption that Vmax has evolved to match contractile frequency is probably overly simplistic, inasmuch as shortening velocity is only one determinant of contractile frequency (8, 69,70). Shortening deactivation and inertial load, for example, are other factors that affect contractile frequency (70). Therefore, Vmax is likely an important, but not exclusive, factor that sets frequency.
Whatever the precise mechanisms involved, the current study demonstrates that the scaling relationship is more universal than previously appreciated, applying not only to terrestrial mammals, but to a wide variety of animals and locomotory mechanisms. In fact, 65% of the variability in Vmax can be accounted for by body mass alone in the group comprised of terrestrial and flying animals (r 2 = 0.65) and this value increases to 78% when experimental temperature is factored into the model. Such a strong correlation might be viewed as surprising, given the wide array of functional demands of different skeletal muscles that are used not only as motors but also as struts, brakes, and springs (25, 34,71). Much of the variability remaining in the data is probably a consequence of these different functional requirements. For example, Rome and colleagues (87) demonstrated that carp possess fast and slow fiber types that differ in shortening velocity by a factor of about three. These different muscle types are coupled to specific biomechanical orientations that either drive slow swimming or fast starts. As such, the fast and slow muscles function as different gears to power specific types of locomotion (87). Likewise, mammalian fibers exhibit several-fold differences in shortening velocity between the fastest and slowest fibers within a given species (Table 1).
Interestingly, even muscles not directly responsible for locomotion can be influenced by locomotory processes. A case in point, the natural frequency of diaphragm muscles in terrestrial mammals is matched with that of stride frequency, reflecting the need to coordinate respiration with movement (4, 103). It would not be surprising to find that shortening velocity in these muscles scales to match locomotory processes, even though the diaphragm is not used for locomotion. Such correlation will tend to reinforce the observed scaling effects, even when the muscles are not directly powering locomotion, but may be acting in other roles such as joint stabilization. The allometries observed for flying and terrestrial animals (Figs. 1, B-D) are particularly interesting when compared with the lack of any scaling in the swimming and nonlocomotory muscles (Fig. 1, E andF, respectively). This juxtaposition suggests that common elements in the physical constraints of weight-bearing animals have had convergent or conserved effects on their locomotory systems.
The muscles used for swimming are conspicuously different from other muscles used for locomotion, with shortening velocities that do not appear to be influenced by body mass. Previous data on the scaling of shortening velocity in swimming muscles have been sparse and equivocal. James et al. (46) found Vmax for fast muscles in Myoxocephalus scorpius scales with a mass exponent of about −0.10 and tail beat frequencies in cod were found to scale with a similar mass exponent of about −0.16 (estimated from Ref.3 assuming fish length = mass0.31). In addition, myofibrillar ATPase activity has been shown to decrease as a function of size in several teleost species (102). In contrast, Curtin and Woledge (24) found the shortening velocity of dogfish muscles to be scale independent. It is tempting to speculate that the apparent scale independence of shortening velocity is related to the effects of neutral buoyancy, but there may also be artifacts of the data that obscure significant trends. For example, the largest swimmers in the data set are fast-swimming pelagic fish that could be expected to possess faster muscles than the other animals. In addition, the body mass range for this group is much narrower than for the flying and terrestrial animals, making significant scale effects more difficult to detect. Nevertheless, the flying and terrestrial muscles still show a significant scaling effect over the more limited mass range of the swimming muscles (P < 0.0001; data not shown). Further work is needed to clarify the pattern for swimming muscles revealed in the current study.
The in vivo data from bird flight muscles also present an intriguing exception to the allometry of other flying and terrestrial muscles. These data were collected using strain gauges implanted in the flight muscles (10, 11, 15, 16, 98, 101) and although the shortening velocities show a scaling effect similar to the other locomotory muscles (mass exponent of −0.13), the regression line is significantly higher than that of the other flying and terrestrial muscles. It is interesting that during flight, the pectoral muscles experience a rapid stretch immediately preceding the contraction producing the downstroke of the wings. This prestretch to 110–125% of resting muscle length is often even greater in magnitude than the change in length experienced during shortening (15, 98, 101). Prestretch of muscle fibers is known to enhance force production (8, 26, 27) and such enhancement has been implicated as an important mechanism for increasing power output during locomotion (8, 101). It may be that this stretch is related to the enhanced shortening velocity as well. A related possibility is that the assumption of power being maximized at V/Vmax of 30% is not be valid for these muscles. Storage of elastic energy in the muscles and tendons may act to amplify power output and these muscles may be operating at a V/Vmax of much greater than 30%. Peplowski and Marsh (80) demonstrated that the power output from tree frogs during jumping was at least seven times higher than predicted from the contractile properties of isolated leg muscles. Their interpretation was that elastic components of the musculoskeletal system were being used to amplify power output. Whatever the reason for the discrepancy between the in vivo bird data and the rest of the values, there is clearly a need for both in vitro and in vivo data to obtain a full understanding of avian flight muscles (86). This need stems from the principle that muscles operate differently in vivo than under experimentally imposed conditions (8, 9, 70, 71). In a recent set of studies, the first comparisons of in vitro and in vivo function of bird flight muscles have become available (10,11). The authors cite preliminary results from in vitro measurements suggesting a V/Vmax of 0.24 and a Vmax of 32 L/s for these flight muscles (10). These values are in good agreement with the estimates used in the current study, but more studies of this type are needed to reveal how the flight muscles of birds may differ from other skeletal muscles.
Many animals change body mass by several orders of magnitude during ontogeny, and the current study suggests that Vmax should decrease as an animal increases in size. Consistent with this prediction, James et al. (46) found that shortening velocity decreased as a function of mass in the swimming muscles of the short-horn sculpin, with shortening velocity scaling as mass−0.10. A detailed protein analysis revealed a shift in the troponin I isoforms in larger animals, with no detectable changes in MHC isoforms. Likewise, myofibrillar ATPase activity in other teleost fish slows as fish increase in size (102), but no information is available about the molecular changes responsible for this slowing. Marsh (69) similarly found that larger lizards possess slower muscles, with Vmax scaling as mass−0.084, but again, no analyses were performed to examine the molecular changes responsible for the adjustments in contractile properties. It is currently unknown how general such changes during growth may be and the mechanisms responsible for these changes remain obscure. Nevertheless, changes in Vmaxlikely require the switching of myofibrillar isoforms during growth and development. A recent study of ontogenic changes in dragonfly flight muscles provides an excellent example of the type of study that might be applied to the questions presented here (68). In these flight muscles, alternative splicing of Tn-T transcripts was found to produce significant changes in wing beat frequency and power output resulting from changes in Ca2+ sensitivity (68). Such fine tuning of myofibrillar proteins during growth in animals exhibiting extremes in mass should provide interesting models of muscular plasticity in the future.
Finally, the current analyses demonstrate that muscle force production is more constant than shortening velocity, but that significant differences do exist, with faster muscles producing slightly greater forces (Fig. 3 A). Although this trend was statistically significant (P < 0.039), the correlation between force production and contractile velocity was weak (r 2 = 0.07) and was likely due to differences in relative myofibrillar volumes. It is expected, for example, that fast glycolytic fibers will possess a relatively higher myofibrillar volume and thus produce greater forces than slow oxidative fibers (62). The slightly higher Po values for the amphibians and mollusks may also be related to such differences. In contrast, the higher force production in some crustacean muscles arises because their long-sarcomered fibers (6–15 μm) possess a higher number of myosin cross bridges per sarcomere (52, 97). The reduced force production related to a lowered duty ratio in some high-frequency muscles (85) does not appear to be a general phenomenon linked to shortening velocity. Overall, the relative constancy of force per cross-sectional area underscores the conserved nature of the myosin molecule. It seems likely that the amount of force produced per myosin cross bridge is a highly conserved trait that became optimized early in the evolution of the myosin molecule.
Although skeletal muscle is often touted for its diversity and specialization, in the broadest sense, skeletal muscles are conserved in their contractile properties. Recent molecular analyses demonstrate that sequence divergence is restricted to limited regions of the MHC molecule and that a greater degree of similarity exists among orthologous MHCs (i.e., human MHC I/mouse MHC I) than among paralogous MHCs (i.e., MHCs I, IIa, IIb, IIx within a species) (64,99). These patterns indicate that functional demands are generally more important than phylogenic history in determining myosin structure and function. The results of the current analysis are consistent with this interpretation, as body mass and functional requirements (terrestrial, flying, swimming, or nonlocomotory) are significant predictors of shortening velocity that transcend differences related to phylogeny. These factors have produced convergent or conserved patterns in the contractile properties of skeletal muscles in a diverse group of animals. The principle pattern revealed here is that the influence of body size on the shortening velocity of skeletal muscles is more general than previously recognized. It is not only the muscles of mammals, but muscles from a wide variety of flying and terrestrial animals that conform to this same allometry. This scaling effect is in sharp contrast to the scale independence of shortening velocity in swimming and nonlocomotory muscles. This juxtaposition suggests that weight-bearing animals in particular have been significantly influenced by common constraints during the evolution of a variety of specific locomotory mechanisms. Integrative and comparative approaches have identified several general principles of locomotion and muscle design common to diverse animals (25, 86) and such approaches will continue to play an important role in our understanding of skeletal muscle function.
I express much appreciation to J. Cherry, T. H. Dietz, K. Medler, and D. L. Mykles for critically reviewing and making helpful suggestions about the manuscript.
This work was supported by a National Institutes of Health National Research Service Award (1-F32-AR-08597–01A1) awarded to S. Medler.
Address for reprint requests and other correspondence: S. Medler, Dept. of Biology, Colorado State Univ., Fort Collins, CO 80523 (E-mail:).
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. Section 1734 solely to indicate this fact.
March 22, 2002;10.1152/ajpregu.00689.2001
- Copyright © 2002 the American Physiological Society