Temperature and muscle

Mechanical performance of muscle is greatly influenced by temperature, as are most biological processes. Maximal forces developed by muscles and their rates of force generation, contraction, relaxation and power output are all altered when body temperature varies. As these muscular forces and rate processes underlie behavioural capacities, these may also be thermally dependent. Such factors as maximal locomotor speed and reaction rates may change with temperature and be so slow in the cold that effective escape or pursuit by an animal is curtailed. In animals that are subjected to varying or low body temperatures, we might expect to find adaptations which minimize the thermal dependence of muscle performance.

The influence of temperature on the mechanical performance of vertebrate skeletal muscle is reviewed in this paper, first for one species of lizard and then for vertebrates more generally. Adaptations of muscle performance to temperature are examined, underlying mechanisms of thermal dependence are reviewed and implications for behaviour are discussed.

First, we will examine the influence of temperature on the contractile performance of muscle from a lizard. Lizards are a particularly useful group of animals for such studies as they may naturally experience a wide range of body temperatures daily. Many species also have very high thermal tolerance. Data are presented here (Fig. 1) on the influence of temperature on several aspects of force generation and contraction of skeletal muscle of the lizard Dipsosaurus dorsalis (Marsh & Bennett, 1985). Dipsosaurus inhabits the hot desert regions of California and regulates field body temperatures at approximately 40°C. However, it must maintain the capacity to react and move even when its body temperature is far below these diurnal levels. The reported data were measured on the white portion of the iliofibularis muscle of the hind limb. This preparation is composed almost exclusively of fast glycolytic fibres (Gleeson, Putnam & Bennett, 1980), which constitute the large majority of fibres in all the locomotory muscles of this species (Putnam, Gleeson & Bennett, 1980).

All contractile rate processes in this muscle, both isometric and isotonic, are greatly accelerated by increasing temperature, even up to 44°C. In isometric twitch, the rates of both tension development (measured as the inverse of time-to-peak tension, TPT) and relaxation (measured as the inverse of time to return to 50 % of maximal twitch tension, 1/2 RT) have Q ₁₀ values of 2·36 and 2·82, respectively, between 20 and 30°C. Maximal rate of isometric tetanic tension development (dP₀/dt) has a Q₁₀ of 2·22 over this range. The maximal velocity of shortening (V_max) and maximal power output during isotonic contractions are similarly thermally dependent with Q₁₀ values of 1 ·95 and 2·42, respectively. All these rates have a thermal dependence similar to those of most other biological processes (i.e. Q₁₀ = 2-3, Precht, Christophersen, Hensel & Larcher, 1973; Prosser, 1973).

In contrast, the temperature dependence of force exerted during isometric contraction is substantially different from that of contractile rate processes. Tetanic tension (P₀) has a significant but very low thermal dependence: its thermal ratio (Rio, the ratio of two quantities measured over a 10°C interval, Bennett, 1984) is 1·2 between 20 and 30°C. Maximal P₀ is attained at 40°C. Twitch tension (P_t) is maximal at 15°C, declining at higher temperatures (R₁₀ = 0·62 between 20 and 30°C).

Force and its rate of development thus have very different thermal sensitivities in this lizard. Nearly maximal muscular performance is obtained at normal field activity temperature (40°C): contraction and relaxation rates are rapid and tetanic force is maximal. Twitch tension is the only factor that is not near its maximal value at this temperature. Exposure to low temperatures greatly retards the speed of muscle contraction in Dipsosaurus. This dependence may restrict locomotory responses at low body temperature : the speed of limb movement during burst escape speed in this lizard is limited by the time course of the muscle twitch at 25 °C and below (Marsh & Bennett, 1985). Low temperature does not similarly affect force output. Nearly the same tetanic tension can be produced and twitch tension is even increased.

How representative is the thermal dependence of muscle function in Dipsosaurus ? Q₁₀ and R₁₀ values measured approximately between 20 and 30°C from all other studies available on vertebrate skeletal muscle are given in Fig. 2. Most of these observations were made on amphibian and mammalian muscle. Although the variance in the data is high, due to such factors as the diversity of animals examined, differing fibre type composition of the muscles and different measurement techniques, the pattern is clear. As in the lizard muscle, rate processes in general are thermally dependent with Q₁₀ values of approximately 2, tetanic tension has a very low thermal dependence and twitch tension often decreases over this temperature range.

This general pattern of high thermal dependence of rate processes might be expected to pose problems for poikilothermic organisms. As muscle temperature changes, so do contractile speeds and possibly reaction rates or locomotor ability. Poikilotherms with low body temperatures might not attain maximal performance of which their muscles are capable due to an extrinsic factor, temperature, rather than an intrinsic structural or biochemical limitation. For example, the lizard Gerrhonotus multicarinatus has field active body temperatures of about 25 °C, but maximal rates of muscle contraction and burst escape speed at 35-40°C (Bennett, 1980; Putnam & Bennett, 1982), body temperatures far in excess of those encountered under any natural conditions. Such a pattern of thermal dependence does not appear particularly adaptive. To what extent has there been adjustment to minimize the thermal perturbation of contractile rate processes in animals with low and/or variable body temperatures? Few data on this topic exist, so no confident generalizations are possible, but the studies available suggest some interesting adaptive patterns.

When an individual animal is exposed to a new thermal regime, it often shows compensatory changes in its physiological reactions (acclimation). Biological rate processes are initially altered in accordance with their Q₁₀ upon acute temperature exposure. Over several days or weeks, these rates often return partially or completely to their original levels, even while the new thermal regime is maintained (Precht et al. Prosser, 1973). Several comprehensive studies on the effect of long-term temperature exposure have been done on the mechanical performance of skeletal muscles of anuran amphibians. As illustrated for twitch kinetics (Fig. 3), no acclimation of either force generation (P_t, P₀) or contractile rate (TPT, 1/2 RT, dP₀/dt, V_max,⁠) has been demonstrated (Renaud & Stevens, 1981a,b; Rome, 1983). The initial depression of contraction rates by cold exposure is maintained indefinitely. A similar lack of acclimation occurs in locomotor capacity of these animals (Putnam & Bennett, 1981). In fish, myofibrillar ATPase activity, which should be reflected in V_max (Bárány, 1967), has been shown to acclimate in goldfish (Johnston, 1979) but not in killifish (Sidell, Johnston, Moerland & Goldspink, 1983).

The lack of acclimation of muscle function is a very puzzling result in view of its obvious importance to behavioural capacity and may reflect a constraint on adaptation.

Studies comparing species naturally exposed to different thermal regimes show a different pattern of adjustment of muscle function over evolutionary time. In lizards, species with lower activity temperatures have lower Q₁₀ values for TPT, 1/2 RT and dP₀/dt and faster twitch responses measured at any common temperature (Putnam & Bennett, 1982). Maximization of P_t at preferred thermal levels previously reported (Licht, 1964) has not been confirmed (Putnam & Bennett, 1982). The range of temperatures over which lizard muscles can function is clearly affected by their thermal regimes (Ushakov, 1964; Licht, 1964; Putnam & Bennett, 1982): muscles from more thermophilic species lose contractile ability and undergo irreversible heat damage at substantially higher temperatures than do those of more cryophilic animals. In fish, actomyosin ATPases of species from cold environments have both lower Q₁₀ values and greater activities at any common temperature than those from warm-adapted fish (Fig. 4) (Johnston, Walesby, Davison & Goldspink, 1977; Johnston & Walesby, 1977, 1979). In both fish and lizards, evolutionary adaptation to temperature has evidently proceeded with both a shift (translation) and rotation of the rate-temperature curve.

It should be emphasized, however, that these interspecific differences are only relative. In absolute terms, rate processes are still very thermally dependent even in cold-adapted species.

Are there major differences in the thermal dependence of muscle function in homeotherms and poikilotherms? One might anticipate that the latter would show adaptations to minimize thermal dependence and the former would optimize function over a narrow range of high temperatures. It should be noted, however, that even muscles from homeotherms may undergo major temperature changes depending on ambient conditions and work intensity (e.g. Saltin, Gagge & Stolwijk, 1968).

Some functional differences are apparent between muscle from anuran amphibians and mammals (Bennett, 1984). Anuran muscle develops maximal P_t at 0°C; mammalian peak P_t usually occurs at 20°C. Maximal P_o is maintained at lower temperatures in anurans: P₀ of anuran muscle declines below approximately 15 °C, that of mammalian muscle always decreases below 25 °C (Fig. 5). The time course of a muscle twitch, both TPT and 1/2 RT, has a lower Q₁₀ in anurans than in mammals. Anuran muscle is thus capable of producing maximal tension at lower temperatures than is mammalian muscle and its rate processes are less temperature sensitive, at least in regard to twitch kinetics. However, the distinction between the thermal dependence of saurian and mammalian muscle function is much less clear (Bennett, 1984). Although maximal P_t is developed at lower temperatures in lizard than in mammalian muscle, the thermal dependencies of PQ and contractile rate processes are almost identical in these groups. The anuran-mammalian differences probably reflect adaptation to very different thermal regimes rather than a dichotomy in the functional capacity of muscle from poikilotherms and homeotherms.

It is apparent from the foregoing that some adjustments of vertebrate skeletal muscle function have evolved with respect to temperature. These adaptations in rate processes, however, have not been substantial. Muscle from poikilotherms still has a marked thermal dependence and rate processes are often faster at temperatures above those normally encountered. Lack of acclimation of these properties in individual animals further points to a lack of plasticity in these systems in regard to temperature adjustment. In view of the small number of studies, it should be evident that more comparative work is required to substantiate or alter these conclusions.

The biochemical and functional properties that underlie these patterns of thermal dependence are not well understood. The pronounced thermal dependence of contractile rate processes are expected, as several of the steps underlying contraction and relaxation are enzymatically catalysed and have rates with Qi₀ values of 2·0 or more. For instance, strong thermal dependencies have been demonstrated for the rate of Ca²⁺ release (Blinks, Rüdel & Taylor, 1978; Rail, 1979), the activity of actomyosin ATPase (Bendall, 1964; Hartshorne, Barns, Parker & Fuchs, 1972) and Ca²⁺ uptake by sarcoplasmic reticulum (Yamamoto & Tonomura, 1967; Blinks et al. 1978). However, the correspondence between the thermal dependence of contractile events and isolated enzyme systems associated with them is often poor. For example, although actomyosin ATPase is thought to be an important determinant of V_max, the thermal dependence of the former is much greater than that of the latter (Bárány, 1967). An attempt to determine limiting factors of contractile events by comparisons of thermal dependencies would probably be no more successful than previous attempts to elucidate controlling reactions of biochemical pathways by measurement of Arrhenius activation energies.

Twitch tension is generally maximal at 0—20°C, depending on species and muscle type, and declines at higher temperatures. In some sense, the muscle is not as fully activated by a single stimulus at temperatures above 20°C (Ranatunga, 1977). The ratio of P_t/P₀ approaches 1·0 at low temperatures but is less than 0·5 at higher temperatures (Putnam & Bennett, 1982). One possible explanation for this response is insufficient Ca²⁺ release into fibres at higher temperatures. This is evidently not the case, however, as activation heat, which represents the energetic cost of Ca²⁺ release and removal, is thermally independent, indicating no deficiency of release at higher temperatures (Homsher, Mommaerts, Ricchiuti & Wallner, 1972; Rail, 1979). Additionally, measurements with aequorin indicate high levels of Ca²⁺ within the fibre at higher temperatures (Blinks et al. 1978). The latter measurements also indicate a much shorter time course for the presence of Ca²⁺ within the fibre at higher temperatures and a consequently shorter period of activation. At high temperatures, insufficient time may be available to attain maximal force. Twitch contraction may be looked upon as a competition between actomyosin ATPase activity and Ca²⁺ removal by the sarcoplasmic reticulum (Josephson, 1981). These processes evidently have different thermal dependencies, with the latter being more temperature sensitive; the Q₁₀ of 1/2 RT is generally greater than that of TPT measured on the same muscle (Hill, 1951; Walker, 1960; Bennett, 1984). Thus at low temperatures, greater force can be attained by the contractile proteins, even though their intrinsic contractile velocity is also slowed, before deactivation processes become effective. This explanation of the thermal dependence of P_t, while attractive, has not been critically tested.

Tetanic tension generally has a low but significant thermal dependence below 15—25 °C, depending on species (Fig. 5). As the muscle is supposedly completely activated by repetitive stimuli, one might expect P₀ to be thermally independent over the entire temperature range. One explanation for lower P₀ at low temperatures is a smaller number of cross-bridge attachments at those temperatures. However, measurements of instantaneous stiffness, which is dependent upon the number of cross-bridges attached, do not support this hypothesis (Ford, Huxley & Simmons, 1977; Kuhn et al. 1979; Bressler, 1981). Instantaneous stiffness is either thermally independent or has a different thermal dependence from that of P₀. Consequently, the number of cross-bridges attached or a thermally-dependent bridge cycling does not appear to account for the observed pattern. Bressler (1981) suggests that increasing temperature may increase the tension per cross-bridge during tetanus.

It should be clear that further work is required to explain the bases of the thermal dependence of muscle function. Investigations exploiting thermal dependence may well lead to further hypotheses and insights into the nature of the contractile process itself (e.g. Mittenthal, 1975, on the distance of cross-bridge movement). Muscle from animals and/or animal groups with very different thermal histories could be particularly useful in these studies. For example, the different response of PQ to temperature in anuran and mammalian muscle (Fig. 5) may be helpful to understanding patterns of cross-bridge tension generation and cycling during tetany. Or, the differing thermal dependencies of myofibrillar ATPase from animals adapted to different temperatures (e.g. Fig. 4) may help elucidate their role in determining intrinsic shortening velocities.

Is the thermal dependence of muscle function reflected in animal behavioural capacities? On the basis of experiments on isolated muscle, one would expect performance capacity for behaviour involving rates of contraction (e.g. running speed) to improve with increasing temperature. One would also expect performance involving maximal force generation by an animal to be relatively temperature independent.

Physical performance involving rate-dependent factors, such as maximal power output, does improve significantly with increasing muscle temperature (e.g. Asmussen & Bøje, 1945; Binkhorst, Hoofd & Vissers, 1977; Bergh & Ekblom, 1979). Consequently, ‘warming up’, in the literal sense, does have a positive effect on performance speed, and animals with higher body temperatures do in fact have greater maximal speeds (Webb, 1978; Bennett, 1980; Putnam & Bennett, 1981). However, the thermal dependence of this performance is distinctly lower than that of rate processes in isolated muscle. For example, in the lizard Dipsosaurus, maximal running velocity and limb cycling frequency have Q₁₀ values of 1·3-1·4 from 25 to 40 °C, while V_max and have Q₁₀ values of 1·7 and 2·0, respectively (Fig. 6) (Marsh & Bennett, 1985). Similarly low Q₁₀ values of 1·0-1·6 have been reported in other studies on rate-dependent performance cited above. This lower thermal dependence of behavioural performance compared to that of isolated muscle function could be attributable to several factors, including storage of energy in elastic structures of low thermal sensitivity (Marsh & Bennett, 1985). However, its basis is unknown at present.

Maximal force generation by muscles in vivo is almost independent of muscle temperature from 25 to 40°C (Binkhorst et al. 1977; Bergh & Ekblom, 1979; Petrofsky, Burse & Lind, 1981). This pattern accords very well with the observed thermal independence of P₀ over this temperature range. Endurance, measured as the time of maintenance of constant force, is maximal at approximately 30°C in both in vivo performance (Clarke, Helion & Lind, 1958; Petrofsky & Lind, 1969; Edwards et al. 1972) and in isolated muscle (Petrofsky & Lind, 1981; Segal & Faulkner, 1982). Endurance declines at both higher and lower temperatures.

As the foregoing data indicate, behavioural performance involving both speed and force generation reflects the underlying patterns of thermal dependence of muscle function. Higher muscle temperatures may be expected to result in improved rate performance in vivo, although maximal force application may be little affected and endurance may decline.

Financial support for this work was provided by NSF Grant PCM 81-02331. I thank Roger Seymour for his helpful comments on the manuscript.