Effects of Physical Training the Metabolism of Skeletal Muscle JAN HENRIKSSON, MD, PHD

With moderate training (30-60 min daily at 70-80% of Vo2 max, 3 - 5 times weekly), the trained muscles display a 40—50% increase in the content of mitochondrial oxidative enzymes. Concomitantly, the total number of muscle capillaries may increase by 50%, whereas the content of glycolytic enzymes is not, or only marginally, affected. The oxidative enzyme increase, which occurs over 6 - 8 wk, is lost in 4 - 6 wk if training is stopped. This loss occurs faster than the decrease in muscle capillarization and in the whole-body Vo2 max. Trained muscles of athletes have 3 - 4 times higher oxidative enzyme levels and two- to threefold more capillaries per muscle fiber than untrained muscle. Extensive endurance training results in an enhanced percentage of slow-twitch fibers, but the time course of this change is not known. More extensive changes are observed in chronically stimulated rabbit muscle. In this case, enzymes of oxidation display large increases (6- to 12-fold), whereas there is a decrease of 70—90% in enzymes of glycolysis, glycogenolysis, gluconeogenesis, and high-energy phosphate transfer. There is a normal training response in mitochondrial enzyme activities in individuals with insulin-dependent and noninsulin-dependent diabetes, but the ability to form new skeletal muscle capillaries in response to physical training may be deficient in insulin-dependent diabetes. Training-induced changes in the metabolic character of skeletal muscle leads to an increased reliance on fat metabolism during exercise, with a lowered blood lactate concentration and a sparing of muscle glycogen.

S

everal important metabolic consequences of physical training—such as the increased reliance on fat metabolism during exercise with a lowered blood lactate concentration and a sparing of muscle glycogen, as well as an increased peripheral insulin sensitivity— are the result of changes in the metabolic character of skeletal muscle. Skeletal muscle tissue represents a large portion of the body's total metabolism and, consequently, these training-induced

changes will have important consequences for the body's total homeostasis, especially during physical activity. The capacity of skeletal muscle cells for adaptation to changes in metabolic demand has been shown to be quite remarkable, and it is a well-known fact that endurance training induces marked adaptive changes in several structural components and metabolic variables in the engaged skeletal muscles. Among the observed changes with different endurance train-

FROM THE DEPARTMENT OF PHYSIOLOGY III, KAROLINSKA INSTITUTE, STOCKHOLM, SWEDEN. ADDRESS CORRESPONDENCE AND REPRINT REQUESTS TO JAN HENRIKSSON, MD, PHD, DEPARTMENT OF PHYSIOLOGY III, KAROLINSKA INSTITUTE, BOX 5626, S-114

DIABETES CARE, VOLUME 15, SUPPLEMENT 4, NOVEMBER

86, STOCKHOLM, SWEDEN.

1992

ing regimens are those involving the muscle's content of metabolic enzymes, sensitivity to hormones, and composition of the contracting filaments. Other adaptations affect membrane transport processes and the muscular capillary network. These changes act with adaptations in the cardiovascular system, autonomic nervous system, and endocrine system to increase an individual's physical working capacity, and to allow a given exercise intensity to be achieved with less strain and metabolic disturbance. This article is devoted to a closer look at some aspects of muscular adaptation to endurance training.

MAXIMAL ADAPTABILITY OF SKELETAL MUSCLE: EFFECTS OF CHRONIC ELECTRICAL STIMULATION — The general nature of the changes induced in skeletal muscle in response to increased physical activity is illustrated by the changes observed in chronically stimulated rabbit muscle ( 1 3). This model allows the study of the response of a normally rather inactive muscle (the rabbit anterior tibial muscle) to a near maximal degree of activity, with virtually no discomfort to the animal. Although there are important points of difference between the complex patterns of impulse activity imposed on an exercising muscle and the much more extensive, but comparatively simple, pattern used in stimulation experiments, there are more than enough similarities in their effects to justify the view that the same basic mechanisms are involved. It appears that endurance exercise and chronic stimulation differ mainly in degree: the properties that change in response to exercise are also those that change at an early stage of stimulation; the properties that are resistant to change under exercise conditions change only after prolonged stimulation. A knowledge of this hierarchy of stability in the properties of skeletal muscle in response to changing functional demands provides a logical framework for under-

1701

Metabolic effects of physical training

IOOO standing the effects of different forms of training or inactivity on the metabolic character of skeletal muscle. During anesthesia, a miniature electronic stimulator is implanted under aseptic conditions to subject the anterior tibial (TA) muscle (when activated) to chronic stimulation via the common peroneal nerve. The stimulator is activated noninvasively by means of an electronic flash-gun after the rabbit has been alweeks lowed to recover from the operation. Figure 1—Enzyme changes induced by When stimulating with a continuous chronic electrical muscle stimulation. Rabbit antrain of pulses at a frequency of 10 Hz, as terior tibial muscle was stimulated at 10 im- Figure 2—Enzyme changes induced by used in most studies, there is a very small pulses/sec, 24 h/day,for different periods of time chronic electrical muscle stimulation. 3-Keamplitude oscillation of the hindpaw, (3 days-10 wk). Changes depicted in the muscle toacid-coenzyme A (CoA) transferase is the first but no observable effect on the use of the content of three oxidative and two glycolytic en- enzyme in ketoacid catabolism; glycogen phoslimb in posture control, locomotion, or zymes. Succinate dehydrogena.se (SDH), citratephotylase and phosphoglucomutase represent synthase, and malate dehydrogenase (MDH) areglycolysis; and glycogen synthase is involved in general well-being of the animal. enzymes in the citric acid cycle, lactate dehydro-the glycogen biosynthetic pathway. The main The rabbit TA muscle is a pregenase (LDH), and P-fructokinase (6-phospho- Junction of hexokinase is to make glucose—taken dominantly fast muscle, containing not fructokinase) are involved in glycolysis. Valueforfrom blood—available to muscle cell by channelmore than 6% slow-twitch fibers. The unstimulated control muscles has been set at ing it into the glycolytic pathway. See Fig. 1 chronic stimulation induces a striking fi100%. From Henriksson et al. (53). © by the legend. From Henriksson et al. (53). © by the ber-type transformation so that after American Journal of Physiology. American Journal of Physiology. stimulation durations of 5 - 6 wk or more, the TA muscle contains slowtwitch fibers only. Simultaneously, the normally very fatiguable TA muscle be- crease, 11-fold at its peak, was linear way that is elevated by chronic stimulacomes highly fatigue resistant. It is likely during the first 2 wk, after which a pla- tion; all other glycolytic and glycothat the increased endurance is mainly a teau was reached and followed by a sec- genolytic enzymes display large result of the pronounced enzyme and ondary decrease. Hexokinase II, the prin- decreases (Figs. 1 and 2). This finding microcirculatory adaptations induced by cipal isoenzyme of skeletal muscle, has indicates that glucose phosphorylation the chronic stimulation, but the fiber- been found to be very sensitive to also may be an important step in the type transformation also may be of im- changes in metabolic conditions. Katzen control of skeletal muscle glucose use. portance. For a more detailed descrip- et al. (8) have reported that the low hexo- The relative imbalance between hexokition of effects of chronic stimulation on kinase II activity in diabetic rat muscle nase and other enzymes of glycolysis is muscles, see reviews by Salmons and was restored to normal 2 h after insulin the probable explanation of the increased Henriksson (1), Jolesz and Sreter (2), administration. Very rapid increase in concentration of glycolytic intermediates and Pette (3). The general nature and hexokinase during the initial phase of noted in these muscles (10). magnitude of the enzymatic changes in- stimulation might be interpreted as indiEnzymes of the citric acid cycle, duced in these models have been de- cating a high demand for blood-borne fatty acid fJ- oxidation, and amino acid scribed in a series of publications by glucose to support a higher glycolytic aminotransferases also display large inPette et al. (4-6). rate during the early weeks of stimula- creases on stimulation, but the increases tion before twitch speed and energy con- occur somewhat later. In our investigasumption have decreased, and the mus- tions, maximal changes (6- to 12-fold) Enzyme adaptation cle has become better adapted to were reached after 2 - 5 wk. Thereafter, The stimulation-induced enzyme combustion of both glucose and fatty ac- the concentrations of some of the enchanges are summarized in Figs. 1 and 2. ids. In line with this interpretation, Maier zymes decreased somewhat before stabiOf all enzymes analyzed, hexokinase and Pette (9) showed severe glycogen lizing at a lower level (Fig. 1). It is not showed the most rapid response to chronic stimulation. This is in accord depletion in all fibers after 2 h of stimu- known whether this two-phase pattern with previous studies, e.g., Weber and lation. It is noteworthy that hexokinase is of change is specific for chronic stimulaPette (7). In our investigations, this in- the only enzyme of the glycolytic path- tion or whether it also may occur with f

1702

DIABETES CARE, VOLUME 15,

SUPPLEMENT 4, NOVEMBER

1992

Henriksson

certain endurance training programs. However, to date, there has been no report that this may occur in response to physical training in man or any other species. The physiological meaning of the biphasic changes of some of the enzymes is not known, although the similarity in time course with the fiber-type transformation from fast-twitch to slowtwitch fibers may point to a causal link to the reported difference in chemical energetics between fast-twitch and slowtwitch muscles (42). In contrast to the marked increase in the oxidative enzymes and amino acid aminotransferases, there was a decrease of 70-90% in enzymes that characterize the predominant metabolism of the normal, unstimulated TA muscle, namely enzymes of glycolysis, glycogenolysis, gluconeogenesis, and high-energy phosphate transfer. On discontinuing stimulation, the enzymes whose activities had increased again declined toward the initial level, at first rapidly and later more slowly, and had returned to normal after 5 - 6 wk. This was also true of the glycolytic enzymes, which increased when stimulation was stopped; but, with these enzymes, normalization occurred more rectilinearly (11). Figure 3 shows that the normally wide variation in enzyme content among different fibers in the same muscle, and even between those of the same type, is reduced on chronic stimulation. No corresponding information is available about the effect of endurance training, although some information can be found in Chi et al. (12). An interesting question arises: Are the levels of some enzymes less activity-dependent and instead more related to the myofibrillar ATPase type of the muscle fiber than others? We have explored this question through enzyme measurements in single-fiber samples dissected from freeze-dried transverse cryostat sections. This allowed direct correlation of enzyme levels with fiber type, determined by myofibrillar ATPase staining of adjacent sections. From these

DIABETES CARE, VOLUME 15,

SUPPLEMENT 4,

other enzymes appear to change according to rules of their own. It was concluded that, although the pattern of use is a major determinant of a fiber's metabolic enzyme profile, an effect is also exerted by the specific myosin isoform complement of the fiber.

Other stimulation-induced changes There is a doubling of the number of FBPase blood capillaries per unit muscle crosssectional area, thus greatly improving the Figure 3—Enzyme changes in single skeletal muscle's blood supply (14). The time muscle fibers induced by chronic electrical muscourse of this change has not been studcle stimulation. Anterior tibial muscle of the rabied in detail, but preliminary data indibit was stimulated as described in Fig. 1 legend. cate that it is roughly similar to that of Single fibers were isolated by microdissection the oxidative enzymes. Concomitant from muscles stimulated for different periods of time (2, 3, 5, and 8 wk) and from unstimulated with these changes there is, as previously control muscles. Fibers were subsequently ana-mentioned, a dramatic improvement in lyzed for two enzymes, citrate synthase, as a the endurance of the muscle. In our inmeasure of the fiber's oxidative capacity, and vestigations, this variable has been meafructose bisphosphatase, as a measure of its gly- sured as an index: the remaining muscle colytic capacity. Citrate synthase is a member offorce following a 5-min protocol of inthe citric acid cycle and fructose bisphosphatasetense muscle stimulation divided by the (FBPase) catalyzes the reversal of the 6-phos- muscle force exerted during the first few phofructokinase reaction in glycolysis. As evidentcontractions. This index increases from a from the figure, all fibers in a normal unstimu- normal value of 0.5 to 1.0 in muscles lated (control) muscle have a high content of that have been continuously stimulated glycolytic enzymes, whereas the content of oxi- for 6 wk. Following discontinuation of dative enzymes varies 10-fold. Chronic stimulachronic stimulation, the fatiguability tion induces a high oxidative capacity in all again increases, the time needed for norfibers, whereas the gfycolytic capacity decreases malization (5-6 wk) being similar to to low levels. Figures are moles (citrate synthase) that of the metabolic enzymes and the or millimoles (FBPase) per kg dry weight per capillary supply. An interesting general hour at 20°C. From Chi et al. (13). © by the observation regarding these chronic American Journal of Physiology. stimulation experiments is that the different biochemical and morphological adaptations to chronic stimulation fit into a "first-in, last-out" pattern for the data (13), it was tempting to postulate response to stimulation and recovery. that: 1) many enzymes of glycogenolysis This means that the earlier the stage at and some of oxidation change synchrowhich a parameter changes during the nously with the changes in myofibrillar course of stimulation, the later the stage ATPase fiber-type and that the concenit returns to control levels during recovtrations of these enzymes are adjusted to ery. For further information on this oblevels appropriate to the new fiber type; servation, see Brown et al. (11). 2) many enzymes of oxidative metabolism increase without regard to fiber type This summary of what is likely to and without synchronization with fiber- be the maximal activity-induced adapttype change, but perhaps in synchrony ability of skeletal muscle might serve as a with each other, to levels that may be far background to a description of the effect higher than those present in control fi- of endurance training on skeletal muscle bers of the same ATPase type; and 3) still characteristics.

NOVEMBER

200

1992

300

1703

Metabolic effects of physical training

EFFECTS OF ENDURANCE TRAINING ON HUMAN MUSCLE Background The first reports of effects of endurance training on metabolic enzymes in skeletal muscle (rat) came from Russian investigators in the 1950s (15), but a detailed investigation of these changes was first made by Holloszy et al. (16). The capillary network of rat skeletal muscle influenced by training already had been shown in 1937 by Petren et al. (17). The first human studies on muscle metabolic enzymes were published around 1970 (18,19). During the 1970s and 1980s, improved methodology allowed more detailed studies on humans and other species. For the human studies, small muscle biopsy specimens (20-100 mg) were obtained, usually from the thigh muscle, but also from other muscles like the gastrocnemius, deltoid, and triceps of the upper arm. In these studies, different groups of individuals were compared (e.g., untrained persons vs. athletes in different sports) or, alternatively, a group of previously untrained individuals was studied repeatedly with muscle biopsies taken during a training period. In spite of the fact that different parts of the same muscle may often differ with regard to fiber-type composition, capillary density, and enzyme content, the muscle biopsy technique has proved to be surprisingly useful for these studies. With this technique, even smaller changes may be detected (such as a change in enzyme content or capillary density of 15-20%) when a group of 5 - 6 subjects is studied with single biopsies before and after training. Generally, however, the biopsy technique is not sensitive enough to allow conclusions to be drawn from the analysis of a single sample. With several samples from the same muscle, the methodological error is markedly reduced.

Changes in oxidative enzymes An illustration of the effects of endurance training on human muscle is found in

1704

3x

1.5x 0.7x

control subjects

immobil.

training (months)

athletes

before after chron. stim. (rabbit)

Figure 4—Influence of physical fitness level on skeletal muscle oxidative capacity, measured as the content of citrate synihase. Muscle tissue from normal sedentary individuals has been compared with muscle subjected to encasement in plaster after injury or to 2-3 mo of moderate endurance training and values recorded in topclass cyclists and long-distance runners. As a further comparison, corresponding values for rabbit anterior tibial muscle before and after 3-5 wk of chronic electrical stimulation are indicated to the right. (Human data have been generously placed at my disposal by Dr. Eva Jansson, Department of Clinical Physiology, Karolinska Hospital, Stockholm; results regarding chronic stimulation are from Henriksson et al. [53]. © by the American Journal of Physiology-)

the observed differences between endurance athletes and untrained individuals (Fig. 4). In oxidative enzymes (i.e., enzymes of fatty acid oxidation, citric acid cycle, and respiratory chain), the levels are approximately 3 - 4 times higher in the trained thigh muscle of the athletes than in the thigh muscle of untrained individuals (20; Fig. 4). With total inactivity, such as in a muscle encased in plaster after an injury, the oxidative enzyme concentrations decrease to 7 0 75% of the "untrained level." It can be speculated that lower levels than this would not be compatible with survival of the muscle cell. The contents of muscle oxidative enzymes in athletes is thus approximately four- to fivefold higher than the lowest values observed in inactive

DIABETES

muscle, but it may be supposed that a very long training time would be required for an individual to cover this whole range. A comparison with chronically stimulated rabbit muscle (Fig. 4) reveals that muscles of endurance athletes have ~40% lower levels of oxidative enzymes than these chronically stimulated muscles. The difference with fat oxidation enzymes is somewhat greater. Ignoring possible differences between the rabbit and humans, this result may indicate that the trained muscles of our best endurance athletes have an oxidative capacity that is half to two thirds of the theoretically attainable maximal level. Another important question concerns the magnitude of the enzyme changes that can be attained with a few weeks or months of more moderate endurance training regimens. Here, information is available from a large number of investigations in which different research groups have studied the effects of 2 - 3 mo of training on the oxidative enzyme content of leg or arm muscles. These studies have usually involved bouts of 30-60 min of exercise at intensities corresponding to 70 — 80% of Vo2 max, 3 - 5 times/wk. With a group of previously untrained individuals, the general finding is —40-50% increase in the content of oxidative enzymes in the trained muscles (Fig. 5). This increase occurs gradually over 6 - 8 wk, with the most rapid change taking place during the first 3 wk of training (21).

Changes in glycolytic enzymes The muscle cell's content of glycolytic enzymes is not, or only marginally, affected by endurance training programs of 2 - 6 mo duration. The content of glycolytic enzymes is normally low in the skeletal muscles of endurance athletes, but this finding is entirely explained by the large percentage of slow-twitch fibers in their muscles. The content of glycolytic enzymes in this fiber type is normally only half of that in the fast-twitch fibers. The mean glycolytic enzyme level of endurance athletes' slow-twitch or fast-

CARE, VOLUME

15,

SUPPLEMENT

4,

NOVEMBER

1992

Henriksson

ing, 2 mo of training at high submaximal exercise intensities being sufficient to in— n — vatt.lat.SOHactivity crease the total number of muscle capil-•A— cytochromt oildat* activity laries by 50% (25; Fig. 6). The difference between endurance athletes and untrained individuals with respect to the capillary count per muscle fiber Geg muscles) has been found to be two- to threefold (21). There is a lack of information about to what extent capillary neoformation is dependent on training intensity and duration. Less intense training regimens often are known to result in oxidative enzyme increases without any change in capillarization. When stains for myofibrillar ATPase have been used as the basis for fiber-type classification, most longituditnd of 7 8 training ' weeks nal studies in humans have failed to (10-Kw»»kt) demonstrate an interconversion of fiber TRAINING OE-TRAINING types (i.e., fast-twitch to slow-twitch) in Figure 5—Effect of endurance training on content of oxidative enzymes in human skeletal muscle. response to endurance training. The staA group of previously untrained subjects trained for 10-14 wk on bicycle ergometers (40 min/day, ble nature of a muscle's fiber-type com4 days/wk; the rate of work corresponded to 80% ofVo2 max). Biopsy samples were obtained from is further illustrated by the rethigh muscle at different intervals during the training period, and 2, 4, and 6 wk after cessationposition of training. Muscle samples were analyzed for oxidative enzymes sucdnate dekydrogenase (SDH) (ofsults the of chronic stimulation studies in citric acid cycle) and cytochrome c oxidase (the last enzyme of the respiratory chain). In addition,rabbits. the Although in this situation there subjects' Vo2 max was determined using the Douglas bag technique. It is noteworthy that, in the is a gradual and complete replacement of posttraining period, whole-body Vo2 max was maintained significantly longer than muscle oxidative fast-twitch fibers by slow-twitch ones, enzyme content. From Henriksson et al. (54). © by Ada Physiologica Scandinavica. quite long periods of chronic stimulation are required. The fiber-type changes are also the first to revert to normal when twitch muscle fibers thus, has been found levels in human muscle (12,19). It is our stimulation is discontinued (11). On the to be normal, or even slightly enhanced experience, however, that changes in basis of these findings, the high percent(12,22). This finding is in accord with hexokinase with endurance training in age of slow-twitch (type 1) fibers in enwhat has been observed during chronic humans are of a smaller magnitude than durance athletes and the opposite findstimulation, when there is a complete fi- those of the mitochondrial oxidative en- ing in sprinters therefore, have been ber-type transformation from fast-twitch zymes (23,24). This is different from ascribed to genetic factors (26). Endurglycolytic (type 2b) to slow-twitch (type 1) chronic stimulation, where the increase ance training is known, however, to lead fibers. In this situation, the glycolytic en- in hexokinase exceeded that of most of to a complete type transformation within zyme content of the muscle is decreased to the oxidative enzymes (Figs. 1 and 2). the fast-twitch (type 2) fibers from type 20% of the initial level, a decrease that The difference may be related to the fact 2b to type 2a (27,28). The concept that endurance precisely reflects the large difference in gly- that extracellular glucose may constitute, colytic potential between fast-twitch glyco- in relative terms, a more important sub- training does not change the relative oclytic and slow-twitch fibers in the rabbit strate when a small muscle group is ac- currence of fast- and slow-twitch fibers has been challenged in recent years. It (13). Considering the results with chronic tivated than in whole-body exercise. has been shown, for example, that: J) stimulation, hexokinase would be exendurance training of long duration pected to behave in a different way than the other glycolytic enzymes. In fact, as Changes in capillarization and fiber leads to the appearance offibersintermediate between fast- and slow-twitch (Fig. has been shown in rats and guinea pigs types (21), it has been reported that endurance Skeletal muscle capillarization in man is 7); 2) the muscles of the dominant leg in training results in increased hexokinase rapidly enhanced after endurance train- different types of athletes, such as bad~—O—"• mailmal oiygtn upti

DIABETES CARE, VOLUME 15, SUPPLEMENT 4, NOVEMBER 1992

1705

Metabolic effects of physical training

REGRESSION OF THE TRAINING-INDUCED ADAPTATION AFTER DISCONTINUATION OF \

\. ; * v

TRAINING— An increase in the oxidative capacity of a muscle induced by 2 mo of endurance training is lost in 4 - 6 wk if the training is stopped (Fig. 5). This loss of muscle oxidative enzymes occurs faster than the decrease in muscle capillarization (Fig. 8) and in the wholebody Vo2 max (Fig. 5). The time course of the decrease in muscle oxidative enzyme content following cessation of

before training

after

training

30

20

10

Figure 6—Typical ejfect of 2 mo of endurance training (identical with that described in Fig. 5) on capillary density in human thigh muscle. Capillaries are seen in the upper stain as dark spots on the boundary between muscle fibers (amylase-Periodic acid-Schiff staining). Histochemical type of the muscle fibers is indicated by two lower stains. These are myofibrillar ATPase stains pretreated at different pHs (the middle stain at pH 4.3 and the lower one at pH 4.6). Slow-twitch (type 1) fibers appear dark in both stains, whereas the fast-twitch oxidative glycofytic (type Ua) fibers are light. Fast-twitch glycofytic (type lib) fibers appear light in the middle, but dark in the lower stain. From Andersen and Henriksson (25). © by the Journal of Physiology.

minton players, contain a significantly increased percentage of slow-twitch fibers; and 3) in several studies of detraining the percentage of fast-twitch muscle fibers increases. It is therefore reasonable to conclude that extensive endurance training will result in an enhanced percentage of slow-twitch fibers. The extent

1706

to which this might occur still remains to be demonstrated. The probable reasons that the fiber-type transformation was not seen in the early studies are that: 1) these were of too short duration and 2) the muscles investigated were postural muscles and therefore relatively trained in the pretraining state (29).

right

Figure 7—Response of human triceps brachii muscle to extensive endurance training: changes in fiber types. Seven subjects skied with sledges 500 miles over a period of 36 days. Muscle biopsies were taken in the right triceps brachii before training, and in the right and left triceps brachii after training. Serial transverse sections of muscle samples were stained for myofibrillar ATPase activity. Individual percentages of myofibrillar ATPase intermediate fibers are depicted (type IB and 11C). Based on available information about the nature of these fibers (29), the result is interpreted as a sign of ongoing fibertype transformation. From Schantz and Henriksson (55). © by Muscle & Nerve.

DIABETES CARE, VOLUME 15,

SUPPLEMENT 4,

NOVEMBER

1992

Henriksson

i. triceps brochii »cop K fibre"' •CC (type I ) "CC (type HA)

training, changes had occurred in the normal impulse pattern of fast motoneurons in the spinal cord. TRAINING-INDUCED ENZYME AND CAPILLARY CHANGES IN INSULIN-DEPENDENT (TYPE I) DIABETIC SUBJECTS

Enzyme adaptation Available data suggest that there is a normal training response in mitochondrial Figure 8—Response of the human triceps enzyme activities in individuals with brachii muscle to extensive endurance training:type I diabetes. This was verified in a changes in muscle capillarization. Six subjectsstudy where 10 male type I diabetic subskied with backpacks 930 miles over a period of jects and 10 healthy male control sub8 wk. Muscle biopsies were obtained from triceps jects performed physical training for 8 brachii muscle before and during the training wk with three 45-min sessions each period, and on two occasions following the disweek (24). The training consisted of jogcontinuation of training. Serial transverse secging and running, with the main emphations of the muscle samples were stained for sis on endurance, and was carefully suvisualization of muscle capillaries (amylasepervised and monitored so that both Periodic acid-Schiff stain). Changes in the numgroups trained identically. The pretrainber of capillaries per muscle fiber and in the ing Vo max was similar in the two 2 average number of capillaries in contact (CC) with type I (slow-twitch) and type IIA (fast- groups, and the training resulted in an twitch oxidative glycolytic)fibersare depicted. identical 13% increase. The activities in From Schantz et al. (30). © by Clinical Physi- calf muscle biopsies of the mitochondrial enzymes citrate synthase (26-27%, ology. P < 0.01-0.05) and succinate dehydrogenase (24-25%, P < 0.05) increased significantly and similarly in the two groups, whereas training, as expected, training agrees well with that observed did not result in significant changes in the following cessation of chronic stimulaactivities of the glycolytic enzymes tion in the rabbit. In the latter case, howand glyceralde6-phosphofructokinase ever, the restoration of muscle capillarhydephosphate dehydrogenase. Interestization and the restoration of oxidative ingly, differences were observed regarding enzyme content occur simultaneously. There has been only one detailed inves- hexokinase and lactate dehydrogenase betigation of the enzyme changes that take tween the two groups. The pretraining place during the detraining of individu- hexokinase level tended to be lower in the als who have done endurance training diabetic subjects (20%, P < 0.1). A similar for several years (12; very well-trained, finding was reported by Saltin et al. (31), although not top, athletes). It was found whereas Costill et al. (32) observed no that the oxidative capacity of the slow- difference between type I diabetic subjects twitch fibers rapidly decreased with de- and healthy control subjects. As previously training to the level found in untrained stated, Katzen et al. (8) observed that the control subjects. Interestingly, however, low hexokinase activity of diabetic rat the oxidative capacity of the fast-twitch muscle was restored to normal within 2 h fibers maintained an elevated level after insulin administration. It could therethroughout the studied 12-wk period of fore be speculated that the lower hexokidetraining. One theory put forward was nase activity in the skeletal muscle of the that, because of prolonged endurance diabetic subjects could be explained by a

DIABETES CARE, VOLUME 15, SUPPLEMENT 4, NOVEMBER 1992

relative insulin deficiency, resulting in diminished hexokinase synthesis either primarily or secondarily (through a decreased entry of glucose into the muscle cell). Physical training restored hexokinase activity in the diabetic group to a level similar to that found in the control group. Concerning lactate dehydrogenase, the pretraining enzyme activity was 40% higher in the diabetic group (P < 0.05), which is in accord with earlier studies (31). Physical training tended to decrease lactate dehydrogenase activity in the diabetic group, whereas no change occurred in the control group. The current finding of higher levels of lactate dehydrogenase in diabetic skeletal muscle seems to be in line with the observation that diabetic subjects have a greater increase in blood levels and leg release of lactate during exercise compared with healthy control subjects (33,34). It is possible that the mechanism is a decreased activity of pyruvate dehydrogenase (diabetic rat muscle [35,36]), leading to a compensatory increase in the lactate dehydrogenase activity. Capillary changes The results of the cited study furthermore indicated that the ability to form new skeletal muscle capillaries in response to physical training may be deficient in patients with type I diabetes mellitus. Thus, muscle capillarization (the number of capillaries per fiber) increased only in the healthy control subjects (control subjects 14%, P < 0.01, diabetic subjects ±0%). A deficient formation of new capillaries may be an expression of the microangiopathy of this disorder. The mean, as well as minimum, thickness of the capillary basement membranes of skeletal muscle is significantly greater in type I diabetic subjects than in healthy control subjects. Preliminary data (Wallberg-Henriksson et al., unpublished observations) indicate that this difference between diabetic and control subjects is reduced with endurance training. The cited results should not be taken

1707

Metabolic effects of physical training

SDH

RQ ( l e g )

p m o l e s . g-'.min" 1

r

mmoles min 4.0

1.00

Lactate Release

" 1

3.0

NT

*

2.0

, r

0.90

NT

\ N T

1.0

T

T

\

0

30

min

50

rest" 0 i Uptake

10

20

30

Nv

50

60 min

1.0

Figure 9—Metabolic significance of training-induced adaptation of human skeletal muscle. A group of subjects underwent one-leg endurance training on the bicycle ergometer for 6 wk. With one leg well-trained (T) (the level ofsucdnate dehydrogenase [SDH] being 25% higher than in the nontrained leg [NT], left panel), subjects performed two-leg bicycle ergometer exercise at 70% of the Vo2 max, in which both legs performed identically. Energy metabolism of the two legs could be analyzed and compared by means of arterial and venous catheterization and muscle biopsy analyses. Catheterization made possible measurements ofVo2 and Vco2 of both legs separately. Vco2/Vo2 ratio, known as respiratory quotient (RQ), indicates the relative contributions of carbohydrate and fat to oxidative metabolism. Carbohydrate oxidation only would result in an RQ of 1.0, whereas an RQ of 0.7 indicates that fat is the sole source of energy. Middle panel indicates that fat is a more important energy source for the trained leg than for the untrained one (RQ value being significantly lower for the trained leg). Accompanying greater use of carbohydrates in the untrained leg, there is a larger formation and release oflactate (see right panel). In the trained leg, lactate release is low, and during the end of the exercise bout there is even a tendency toward an uptake of lactate from blood. From Henriksson (56). © by the Journal of Physiology.

to indicate that a capillary neoformation cannot be induced by physical training in diabetic patients. In a previous investigation, we observed a 15% (P < 0.01) increase in the number of capillaries per fiber after 4 mo of physical training in a group of nine male type I diabetic patients. No control group was included in this study (23). TRAINING-INDUCED ENZYME AND CAPILLARY CHANGES IN NON-INSULIN-DEPENDENT (TYPE II) DIABETIC SUBJECTS— Similar to healthy control subjects and type I diabetes patients, type II diabetic individuals engaged in physical exercise programs demonstrate an increase in the activity of mitochondrial oxidative enzymes in skeletal muscle, whereas the activity of glycolytic enzymes remain unchanged (37-39). In

1708

TRAINING-INDUCED CHANGES IN INSULIN SENSITIVITY— Another important effect of physical training is an increased sensitivity to insulin with respect to peripheral glucose uptake. Skeletal muscle appears to be the major site of the increased peripheral insulin sensitivity associated with physical training (40,41). The training effect on tissue insulin sensitivity is lost rapidly after the termination of the physical training period, but a single exercise session can restore the increased insulin sensitivity. Therefore, a large part of the "training effect" is really a residual effect of the last bout of exercise (42,43). Types I and II diabetes mellitus are characterized by peripheral insulin resistance in skeletal muscle (44— 46), a defect that may be counteracted by regular exercise (23,38,40). The muscle insulin resistance in the diabetic state may be explained by defects at multiple sites, and the cellular events responsible for the normalization with physical training remain to be established.

METABOLIC SIGNIFICANCE OF THE TRAINING-INDUCED one study (39), the capillary density of ADAPTATION OF SKELETAL skeletal muscle was found to be un- MUSCLE— After cessation of chronic changed in type II diabetic subjects in electrical stimulation, the restoration of a response to physical training. In contrast, normal muscle oxidative enzyme content Allenberg et al. (37) found that the cap- and capillarization follows a time course illary density of skeletal muscle in a similar to that of the normalization of group of type II diabetic patients was muscle endurance. This indicates that the described adaptations are of imporcomparable with nondiabetics of similar tance for the muscle's capacity to perage. They reported a normal increase in form prolonged exercise. This can be furthe number of capillaries per unit of ther illustrated by an investigation in muscle cross-sectional area in response which a group of subjects underwent to a 10- to 15-wk endurance training one-leg endurance training on a bicycle program in this patient group. These dis- ergometer for 6 wk (Fig. 9). With one leg crepancies with respect to the capacity well-trained (the level of succinate dehyfor capillary neoformation in type II di- drogenase being 25% higher than in the abetics are likely to be caused by differ- untrained leg), the subjects subsequently ences in exercise intensity, duration, and performed two-leg endurance exercise at preexercise fitness level. Vascular lesions, 70% of the Vo2 max, in which both legs secondary to the diabetic state, may ex- performed identically. The energy meplain the lack of response in some pa- tabolism of the two legs could be analyzed and compared by means of arterial tient groups (39).

DIABETES CARE, VOLUME

15,

SUPPLEMENT

4,

NOVEMBER

1992

Henriksson

and venous catheterization and muscle biopsy analysis. As illustrated in Fig. 9, there was a significantly smaller release of lactate from the trained leg than from the untrained one, and a significantly larger percentage of the energy output in the trained leg was derived from fat combustion. It is known from a large number of studies that, at the same absolute exercise intensity, trained individuals rely more on fat as an energy substrate than untrained ones. This is the case despite the fact that, at a given rate of work, the plasma levels of free fatty acids are often lower in endurance-trained subjects (47). The importance of the increased fat reliance after training is illustrated by the finding that humans deplete their muscle glycogen stores less rapidly at a given exercise intensity in the trained than in the untrained state (48). Furthermore, at the same absolute or relative (% of Vo2 max) exercise intensity, trained individuals display less lactate concentrations in blood and muscle than do untrained subjects (49). It is known that, although oxidation of fatty acids can provide essentially all of the energy needed for light exercise, muscle glycogen is indispensable for the performance of prolonged, strenuous exercise (requiring —60% or more of Vo2 max) and that, at these work loads, fatigue coincides with the depletion of muscle glycogen stores (48). Taken together, the available evidence, including the result of the oneleg training study, indicates that the heavier reliance on fat in endurancetrained individuals must be explained to a large extent by local factors within the trained muscle. Such factors may be a larger utilization of intracellular or extracellular adipose tissue stores, and it is likely that the high content of mitochondrial oxidative enzymes is important in this respect. This is supported by the results of a study on the rat in which it could be shown that the amount of glycogen (muscle + liver) remaining after an endurance exercise test on a rodent treadmill was directly proportional to the

DIABETES CARE, VOLUME 15,

SUPPLEMENT 4,

muscle's content of oxidative enzymes (50). Holloszy and Booth (16) have proposed a possible biochemical mechanism whereby a large concentration of oxidative enzymes (i.e., citric acid cycle and fat oxidation enzymes and respiratory chain components) leads to a situation in which a major portion of the energy supply is derived from fat metabolism, a lower rate of lactate formation, and sparing of muscle glycogen during exercise. The training-induced enhancement of muscle capillarization probably contributes to the metabolic adaptation seen in trained muscle. A conceivable mechanism for this effect might involve an augmented muscle supply of oxygen and fatty acids.

POSSIBLE MECHANISMS INDUCING THE ENZYME ADAPTATION TO TRAINING— Enzymes and other protein molecules have a limited life span. They are built up and degraded in a continuous cycle in which the biological half-life of many of the mitochondrial enzymes is —1 wk and that of the glycolytic enzymes one to a few days. Accordingly, the cellular content of a certain enzyme is the result of this balance between synthesis and degradation. A change in the rate of synthesis of enzyme proteins is the most important factor in explaining the enzyme changes resulting from chronic stimulation or training (3, 51,52). An interesting area of research today is the exploration of the biochemical mechanisms underlying the altered rate of enzyme synthesis (i.e., discovering how the information, that there is a need for an increased amount of oxidative enzymes in the muscle cell, is transferred to the genes). Among the suggested mediators are decreases in the concentration of ATP or other highenergy phosphate compounds, a decreased oxygen tension, an increased sympathoadrenal stimulation of the muscle cell, substances released from the motor nerve, and calcium-induced diacylglycerol release with subsequent ac-

NOVEMBER

1992

tivation of protein kinase C. The availability of advanced genetic techniques and improved cell culture systems have led to a renewed interest in this area of research, and this will very likely lead to a better understanding of the mechanisms whereby the skeletal muscle cell adapts to different normal and pathological states.

Acknowledgments—The work in my laboratory was supported by grants from the Swedish Medical Research Council, the Karolinska Institute, the Research Council of the Swedish Sports Federation, and the National Institutes of Health. I am grateful to Ulla Siltberg for help in the preparation of this manuscript.

References 1. Salmons S, Henriksson J: The adaptive response of skeletal muscle to increased use. Muscle Nerve 4:94-105, 1981 2. Jolesz F, Sreter FA: Development, innervation, and activity-pattern induced changes in skeletal muscle. Annu Rev Physiol 43:531-52, 1981 3. Pette D: Activity-induced fast to slow transitions in mammalian muscle. Med Sri Sports Beer 16:517-28, 1984 4. Pette D, Smith ME, Staudte HW, Vrbova G: Effects of long-term electrical stimulation on some contractile and metabolic characteristics of fast rabbit muscles. Pflugers Arch 338:257-72, 1973 5. Pette D, Muller W, Leisner E, Vrbovd G: Time dependent effects on contractile properties, fibre population, myosin light chains and enzymes of energy metabolism in intermittently and continuously stimulated fast twitch muscles of the rabbit. Pflugers Arch 364:103-12, 1976 6. Heilig A, Pette D: Changes induced in the enzyme activity pattern by electrical stimulation of fast-twitch muscle. In Plasticity of Muscle. Pette D, Ed. New

York, de Gruyter, 1980, p. 409-20 7. Weber FE, Pette D: Contractile activity enhances the synthesis of hexokinase II in rat skeletal muscle. FEBS Lett 238:7173, 1988

1709

Metabolic effects of physical training

8. Katzen HW, Soderman DD, Wiley CE: Multiple forms of hexokinase. Activities associated with subcellular paniculate and soluble fractions of normal and streptozotocin diabetic rat tissues. J Biol Chan 245:4081-96, 1970

30. Schantz PG, Henriksson J, Jansson E: Adaptation of human skeletal muscle to endurance training of long duration. Clin Physiol 3:141-51, 1983 31. Saltin B, Houston M, Nygaard E, Graham cle Metabolism During Exercise. Vol. 11. T, Wahren J: Muscle fiber characteristics Pernow B, Saltin B, Eds. New York, Plein healthy men and patients with juve9. Maier A, Pette D: The time course of num, 1971, p. 87-95 nile diabetes. Diabetes 28 (Suppl. 1):93glycogen depletion in single fibers of 99, 1979 chronically stimulated rabbit fast-twitch 20. Saltin B, Rowell LB: Functional adaptamuscle. Pflugers Arch 408:338-42, 1987 tions to physical activity and inactivity. 32. Costill DL, Cleary P, Fink WJ, Foster C, Fed Proc 39:1506-13, 1980 Ivy JL, Witzmann F: Training adapta10. Henriksson J, Salmons S, Chi MM-Y, tions in skeletal muscle of juvenile diaHintz CS, Lowry OH: Chronic stimula- 21. Saltin B, Gollnick PD: Skeletal muscle betics. Diabetes 28:818-22, 1979 tion of mammalian muscle: changes in adaptability: significance for metabolism metabolite concentrations in individual and performance. In Handbook of Physi- 33. Wahren J, Hagenfeldt L, Felig P: fibers. Am] Physiol 255:C543-51, 1988 ology, Section 10. Skeletal Muscle. Peachey Splanchnic and leg exchange, of glucose, LD, Ed. Bethesda, MD, American Physiamino acids and free fatty acids during 11. Brown JMC, Henriksson J, Salmons S: ological Society, 1983, p. 555-631 exercise in diabetes mellitus. J Clin Invest Restoration of fast muscle characteristics 55:1303-14, 1975 following cessation of chronic stimula- 22. Essen-Gustavsson B, Henriksson J: Enzyme levels in pools of microdissected 34. Kemmer FW, Berchtold P, Berger M, tion: physiological, histochemical and Starke A, Cuppers HJ, Gries FA, Zimmerhuman muscle fibers of identified type. metabolic changes during slow-to-fast man H: Exercise-induced fall of blood Adaptive response to exercise. Acta Phystransformation. Proc Royal Soc hand B glucose in insulin-treated diabetics unreiol Scand 120:505-15, 1984 235:321-46, 1989 lated to alteration of insulin mobiliza12. Chi MM-Y, Hintz CS, Coyle EF, Martin 23. Wallberg-Henriksson H, Gunnarsson R, tion. Diabetes 28:1131-37, 1979 Henriksson J, DeFronzo R, Felig P, OstWH 111, ivyJL, Nemeth PM, HolloszyJO, man J, Wahren J: Increased peripheral in- 35. Hagg SA, Taylor SL, Ruderman NB: GluLowry OH: Effects of detraining on encose metabolism in perfused skeletal sulin sensitivity and muscle mitochondrial zymes of energy metabolism in individmuscle. Pyruvate dehydrogenase activity enzymes but unchanged blood glucose ual human muscle fibres. Am J Physiol in starvation, diabetes and exercise. Biocontrol in type 1 diabetics after physical 244:C276-87, 1983 chem] 158:203-10, 1976 training. Diabetes 31:1044-50, 1982 13. Chi MM-Y, Hintz CS, Henriksson J, Salmons S, Hellendahl RP, Park JL, Nem- 24. Wallberg-Henriksson H, Gunnarsson R, 36. Imura M, Takatani O: Short term regulation of mouse pyruvate dehydrogenase Henriksson J, Ostman J, Wahren J: Ineth PM, Lowry OH: Chronic stimulation complex by insulin. Horm Metab Res 13: fluence of physical training on formation of mammalian muscle: enzyme changes 267-72, 1981 of muscle capillaries in type I diabetes. in individual fibres. Am J Physiol 251: 37. Allenberg K, Johansen K, Saltin B: SkelDiabetes 33:851-57, 1984 C633-42, 1986 etal muscle adaptations to physical train14. Brown MD, Cotter MA, Hudlicka O, 25. Andersen P, Henriksson J: Capillary suping in type II (non-insulin-dependent) ply of the quadriceps femoris muscle of Vrbova G: The effects of different patdiabetes mellitus. Acta Med Scand 223: man: adaptive response to exercise. J terns of muscle activity on capillary den365-73, 1988 Physiol 270:677-90, 1977 sity, mechanical properties and structure of slow and fast rabbit muscles. Pflugers 26. Komi PV, Viitasalo JT, Havu M, 38. Krotkiewski M, Lonnroth P, Mandroukas K, Wroblewski Z, Rebuffe-Scrive M, Thorstensson A, KarlssonJ: Physiological Arch EurJ Physiol 361:241-50, 1976 Holm G, Smith U, Bjomtorp P: The efand structural performance capacity: ef15. Kendrick-Jones J, Perry SV: Enzymatic fects of physical training on insulin sefect of heredity. In International Series of adaptation to contractile activity in skelcretion and effectiveness and on glucose Biomechanics, Vol. 1A. Baltimore, MD, etal muscle. Nature (Land) 208:1068, metabolism in obesity and type 2 (nonUniv. Park Press, 1976, p. 118-23 1965 insulin-dependent) diabetes mellitus. 16. HolloszyJO, Booth FW: Biochemical ad- 27. Andersen P, Henriksson J: Training inDiabetologia 28:881-90, 1985 duced changes in the subgroups of huaptations to endurance exercise in musman type II skeletal muscle fibers. Acta 39. Lithell H, Krotkiewski M, Kiens B, cle. Annu Rev Physiol 38:273-91, 1976 Wroblewski Z, Holm G, Stromblad G, Physiol Scand 99:123-25, 1977 17. Petren T, Sjostrand T, Sylven B: Der EinGrimby G, Bjomtorp P: Non-response of fluss des Trainings auf die Haufigkeit der 28. Jansson E, Kaijser L: Muscle adaptation muscle capillary density and lipoproteinto extreme endurance training in man. Capillaren in Herz- und Skeletmuskulalipase activity to regular training in diaActa Physiol Scand 100:315-24, 1977 tur. Arbeitsphysiologie 9:376-86, 1937 betic patients. Diabetes Res 2:17-21, 18. Varnauskas E, Bjomtorp P, Fahlen M, 29. Schantz PG: Plasticity of human skeletal 1985 muscle. Acta Physiol Scand 128 (Suppl. Prerovsky 1, Stenberg J: Effects of physi40. Horton ES: Role and management of ex558, Ch. 3): 14-27, 1986 cal training on exercise blood flow and

1710

enzymatic activity in skeletal muscle. Cardiovasc Res 4:418-22, 1970 19. Morgan TE, Cobb LA, Short FA, Ross R, Gunn DR: Effects of long-term exercise on human muscle mitochondria. In Mus-

DIABETES CARE, VOLUME 15, SUPPLEMENT 4, NOVEMBER

1992

Henriksson

41.

42.

43.

44.

45.

46.

ercise in diabetes mellitus. Diabetes Care 11:201-11, 1988 Koivisto VA, Yki-Jarvinen H, DeFronzo RA: Physical training and insulin sensitivity. Diabetes Metab Rev 1:445-81, 1986 Heath GW, Gavin III JR, HinderliterJM, HagbergJM, Bloomfield SA, HolloszyJO: Effects of exercise and lack of exercise on glucose tolerance and insulin sensitivity. JAppl Physiol 55:512-17, 1983 IvyJL, Young JC, McLane JA, Fell RD, Holloszy JO: Exercise training and glucose uptake by skeletal muscle in rats. J Appl Physiol 55:1393-96, 1983 DeFronzo RA, Hendler R, Simonson D: Insulin resistance is a prominent feature of insulin-dependent diabetes. Diabetes 31:795-801, 1982 Andreasson K, Galuska D, Thome A, Sonnenfeld T, Wallberg-Henriksson H: Decreased insulin-stimulated 3-O-methylglucose transport in in vitro incubated muscle strips from type II diabetic subjects. Acta Physiol Scand 142:255-60, 1991 Bogardus C, Ravussin E, Robbins DC,

DIABETES CARE, VOLUME 15,

SUPPLEMENT 4,

Wolfe RR, Horton ES, Sims AH: Effects Chem 252:416-19, 1977 of physical training and diet therapy on 52. Williams RS, Salmons S, Newsholme EA, carbohydrate metabolism in patients Kaufmann RE, Mellor J: Regulation of with glucose intolerance and non-insunuclear and mitochondrial expression by lin-dependent diabetes mellitus. Diabetes contractile activity in skeletal muscle. J 33:311-18, 1984 Biol Chem 261:376-80, 1986 47. HolloszyJO: Metabolic consequences of 53. Henriksson J, Chi MM-Y, Hintz CS, Young DA, Kaiser KK, Salmons S, Lowry endurance exercise training. In Exercise, OH: Chronic stimulation of mammalian Nutrition and Energy Metabolism. Horton muscle: changes in enzymes of six metES, Terjung RL, Eds. New York, Macabolic pathways. Am] Physiol 20:C614millan, 1988, p. 116-31 32, 1986 48. Saltin B, Karlsson J: Muscle glycogen utilization during work of different intensi- 54. Henriksson J, Reitman JS: Time course of changes in human skeletal muscle succities. In Muscle Metabolism During Exernate dehydrogenase and cytochrome oxcise. Pernow B, Saltin B, Eds. New York, idase activities and maximal oxygen upPlenum, 1971, p. 289-99 take with physical activity and inactivity. 49. Hurley BF, HagbergJM, Allen WK, Seals Acta Physiol Scand 99:91-97, 1977 DR, Young JC, Cuddihee RW, Holloszy JO: Effect of training on blood lactate 55. Schantz P, Henriksson J: Increases in levels during submaximal exercise.JAppl myofibrillar ATPase intermediate human Physiol 56:1260-64, 1984 skeletal muscle fibers in response to endurance training. Muscle Nerve 6:55350. Fitts RH, Booth FW, Winder WW, Hol56, 1983 loszyJO: Skeletal muscle respiratory capacity, endurance and glycogen utiliza- 56. Henriksson J: Training induced adaptation. Am] Physiol 228:1029-33, 1975 tion of skeletal muscle and metabolism during submaximal exercise. J Physiol 51. Booth FW, HolloszyJO: Cytochrome c (Lond) 270:661-75, 1977 turnover in rat skeletal muscle. J Biol

NOVEMBER

1992

1711