Metabolic indicators of fibre recruitment in mammalian muscles during locomotion

The purpose of this paper is to discuss spatial patterns of skeletal muscle fibre activity during locomotory exercise, and to relate these to several metabolic correlates. The data emphasize the presence of metabolic heterogeneity both within and among skeletal muscles with a dependence of metabolic patterns on the fibre activity patterns.

That mammalian skeletal muscles are composed of fibre types with different physiological and metabolic properties is well known. Although there have been a variety of systems suggested for classifying the fibres, most investigators agree they can be divided into three general types (see reviews by Close, 1972; Burke & Edgerton, 1975; Burke, 1981; Saltin & Gollnick, 1983). The nomenclature proposed by Peter et al. (1972), in which the fibres are referred to as fast-twitch-oxidative-glycolytic (FOG), fast-twitch-glycolytic (FG) and slow-twitch-oxidative (SO), will be used in this paper, although recognizing there are some problems in reliably differentiating between FOG and FG fibres using histochemical analysis for mitochondrial oxidative enzymes (Nemeth, Hoffer & Pette, 1979; Gollnick, Parsons & Oakley, 1983). This method (Peter et al. 1972) has the advantage of incorporating both the contractile and metabolic characteristics in the system of nomenclature.

Recognizable patterns of fibre type distribution exist within and among mammalian locomotory muscles (Baldwin et al. 1972; Ariano, Armstrong & Edgerton, 1973; Collates, Edgerton, Smith & Botterman, 1977; Armstrong, 1980; Armstrong, Saubert, Seeherman & Taylor, 1982; Armstrong & Phelps, 1984). Fig. 1 shows the distribution of high-oxidative fibres in the leg muscles of rat. Deep physiological extensor muscles are predominantly or totally composed of SO fibres (e.g. soleus muscle). More peripheral extensor muscles generally contain more fast-twitch fibres (e.g. gastrocnemius muscle), with the SO and FOG fibres primarily situated in the deeper portions of the muscles. FG fibres generally predominate in the most superficial portions of the muscles. This stratification of the fibres is most apparent in the extensor muscle groups of quadrupeds (Ariano et al. 1973 ; Collates et al. 1977 ; Armstrong, 1980), but the pattern can also be perceived in human muscles (Johnson, Polgar, Weightman & Appleton, 1973; Edgerton, Smith & Simpson, 1975; Lexell, Henriksson-Larsen & Sjostrom, 1983).

Not only are the muscle fibre types distributed in regular patterns in the muscles, they also appear to be recruited in reasonably predictable spatial patterns during locomotory exercise. The most direct evidence indicating the general patterns of recruitment of the fibres in conscious animals during locomotion comes from electromyographic (EMG) measurement. These recordings indicate the deep SO extensor muscles, such as soleus, are maximally active during quiet postural standing (Smith, Edgerton, Betts & Collates, 1977; Walmsley, Hodgson & Burke, 1978), as well as during locomotion and jumping (Smith et al. 1977; Walmsley et al. 1978; Gardiner, Gardiner & Edgerton, 1982). On the other hand, more peripheral fast-twitch synergists (i.e. gastrocnemius and plantaris muscles) are progressively activated from walking through running and jumping (Smith et al. 1977 ; Walmsley et al. 1978; Gardiner et al. 1982). Thus, in extensor muscle groups there is a progressive activation of motor units from deep slow muscles to more peripheral fast ones with increasing locomotory speed.

There are also several less direct metabolic indicators of fibre type recruitment patterns during locomotion. One that has commonly been used to infer the patterns of muscle fibre use is to determine glycogen loss in the fibres during a bout of exercise. The assumption is made that if a fibre loses glycogen, it has been active at some time during the exercise (Gollnick et al. 1973; Costill et al. 1973; Armstrong et al. 1974, 1977; Sullivan & Armstrong, 1978). Findings from glycogen loss experiments indicate that during walking, deep SO and FOG fibres are primarily recruited in extensor muscle groups (Fig. 2), and with increasing locomotory speed there is progressive recruitment of more peripheral fibres with lower oxidative capacities (FG fibres) (Armstrong et al. 1974, 1977; Sullivan & Armstrong, 1978). During high speed galloping, FG fibres in the most superficial portions of the muscles are active (Sullivan & Armstrong, 1978).

Similar general conclusions about the spatial patterns of recruitment during locomotion may be drawn from studies of mitochondrial changes in the muscles with training (Dudley, Abraham & Terjung, 1982). As shown in Fig. 3, when rats are trained to run on the treadmill at slow speeds the mitochondrial (cytochrome c concentration) adaptations primarily occur in the deep red muscles and muscle parts (soleus and red vastus lateralis). However, with increasing training speeds, the more peripheral portions (white vastus lateralis) of the extensor muscle groups pro-gressively demonstrate the training effect. Thus, these results can be interpreted as indicating a progressive peripheral recruitment of fibres as the training speed was increased.

We have recently observed that blood flow distribution within and among the muscles in rats appears to be related to the recruitment patterns during postural maintenance and locomotory exercise (Laughlin & Armstrong, 1982, 1983; Armstrong & Laughlin, 1983a, 1984). Thus, when rats are standing on the treadmill, blood flows in the extensor muscles are a function of the SO fibre populations (Laughlin & Armstrong, 1982). When the proportion of a muscle composed of SO fibres is experimentally increased, as in overload hypertrophy (lanuzzo, Gollnick & Armstrong, 1976), the blood flow to the muscle during postural standing is increased proportionately (Fig. 4). Thus, in a standing rat there is a close relationship between the population of SO fibres, which presumably are recruited during postural maintenance (Smith et al. V)TT, Walmsley et al. 1978), and the blood flow to the muscles. However, when the animal begins to move, the increases in blood flow in the muscles over pre-exercise are in direct proportion to the populations of FOG fibres in them (Laughlin & Armstrong, 1982). This is true in rats for speeds from 15 m min⁻¹ (fast walk) to 105 mmin⁻¹ (fast gallop) (Fig. 5). Thus, even at running speeds (i.e. ≥75 m min⁻¹) at which most of the peripheral FG units are recruited in the muscles (Sullivan & Armstrong, 1978), the blood flows are primarily directed to the FOG fibres. In fact, extrapolation of regression curves relating blood flows to fibre type populations suggests blood flow in a hypothetical pure FOG muscle would exceed 600 ml min⁻¹100 g⁻¹ (Armstrong & Laughlin, 1983a). Similar calculations yield values of 250 and 110mlmin⁻¹100g⁻¹ for pure SO and FG muscles, respectively (Armstrong & Laughlin, 1983a). Thus, although the flows are markedly less than those in red muscles, blood flow is significantly elevated in white (FG) muscles at high running speeds (≥60mmin⁻¹) (Laughlin & Armstrong, 1982; Armstrong & Laughlin, 1983a), speeds at which FG fibres are known to be recruited (Sullivan & Armstrong, 1978).

Although the purpose of this paper is to discuss the distribution of flow specifically in muscles, the reader is reminded that the distribution of total cardiac output also changes quite dramatically with increasing exercise intensity (Rowell, 1974). This redistribution in rats is illustrated in Fig. 6. It is apparent that a nearly complete cessation of blood flow to visceral organs may occur under heavy treadmill exercise conditions, which further facilitates delivery of blood flow to the active muscle fibres.

The precision of the matching of blood flow to muscle fibre activity is illustrated by the relative differences in flows in extensor and flexor muscles during quadrupedal locomotion versus swimming in rats (Fig. 7). During pre-exercise and during walking and galloping, flows in ankle extensor muscles are higher than in the antagonistic flexor muscles, but during swimming the flows are higher in the flexor muscles (Laughlin, Mohrman & Armstrong, 1984). In fact, soleus muscle, which always has near maximal blood flow whenever the rat is supporting its body mass in the upright quadrupedal position (Laughlin & Armstrong, 1982), has a significant decrease in flow when the animal begins to swim (Laughlin et al. 1984) (Fig. 8). Also, during the first 5 min of swimming the soleus muscle loses no glycogen, unlike the synergistic plantaris and red gastrocnemius muscles (Armstrong & Laughlin, 19836). These data suggest that soleus muscle is recruited less during the first 5 min of swimming than during postural maintenance or terrestrial locomotion. Flexor muscles are relatively more active during swimming (Gruner & Altman, 1980), so that differences in blood flows in the antagonistic muscle groups during treadmill locomotion and swimming could be predicted.

The close matching between blood flow and muscle fibre activity is often considered to result from the release of vasodilatory substances from the active fibres (for reviews see Skinner, 1975; Sparks, 1980), although some have suggested a more direct link between neural activation of the skeletal muscle fibres and the elevations in blood flow through adjacent vasculature (Antal, 1968; Honig, 1979). The extremely rapid and dramatic elevations in blood flow to muscles that occur at the initiation of exercise (Van Citters & Franklin, 1969; Donald, Rowlands & Ferguson, 1970; Armstrong & Laughlin, 1983a) certainly make it intuitively attractive to suggest such a linkage.

We have recently completed experiments in which we preferentially blocked high-oxidative fibres in the muscles with low dosages of curare during slow treadmill exercise (Armstrong, Vandenakker & Laughlin, 1985). This induced more peripheral recruitment of lower oxidative fibres to maintain tread speed. However, the deep red muscles that were recruited less, and/or were less metabolically active in the curarized rats (as indicated by decreased glycogen loss) during the first minute of exercise, had the same increases in blood flows as the normal animals (Fig. 9), These data suggest that the initial exercise hyperaemia in the red muscles is not dependent upon muscle fibre metabolism. Since the alpha-motor neurones to the deep red motor units in the curarized rats presumably continued to be activated (Henneman & Mendell, 1981), the results support the concept of a direct linkage between the motor neurones and the resistance vessels in the muscles.

Although there is a close association between fibre recruitment patterns and metabolic correlates as discussed above, it is clear that metabolic events alone cannot be used to quantify muscle fibre activity or even to measure the relative degrees of activity in different populations of fibres (Gollnick et al. 1973 ; Armstrong et al. 1974 ; Burke & Edgerton, 1975). Fig. 10 illustrates the progressive elevations in blood flow that occur in rat muscles over time during prolonged, low intensity treadmill exercise (Laughlin & Armstrong, 1983). One of the hypotheses offered to explain this gradual rise was that a progressive recruitment of new motor units occurs as the initially-recruited fibres become fatigued (Laughlin & Armstrong, 1983), as has previously been observed under other conditions (Gollnick et al. 1973; Armstrong et al. 1974). However, in subsequent experiments adductor longus muscle (which had the highest blood flow of all hind-limb muscles after about 1 h of exercise, Fig. 10) showed large decreases in glycogen content in almost all of its fibres during the first 20 min of exercise, with no further loss occurring from 20 to 80 min (Tan & Armstrong, 1984). These data suggest the elevations in blood flow were not related to changes in fibre recruitment.

In summary, several types of metabolic changes within and among muscles may be used to estimate whether the fibres in the particular regions are active during in vivo locomotion. Because of the widely differing oxidative and glycolytic capacities and substrate preferences of the fibres, it is not possible to quantitate their activities, but the use of metabolic indicators does provide insight into the spatial relationships of fibre recruitment patterns in the muscles.

The authors would like to thank Mrs Judy White for the artwork. These studies were supported by NIH Grants AM25472 and HL26963, by AHA-Tulsa Affiliate and by ORU Research Funds.