Exercise Physiology: Chapter 2 PDF

chapter 2 THE NEUROMUSCULAR SYSTEM Alan J. McComas I N the first section, which considers the muscles as the system responsible for gen- erating force and movement, I am heavily indebted to the writings of Sir Michael Foster (50, 51...

chapter 2 THE NEUROMUSCULAR SYSTEM Alan J. McComas I N the first section, which considers the muscles as the system responsible for gen- erating force and movement, I am heavily indebted to the writings of Sir Michael Foster (50, 51), the first professor of physiology at the University of Cambridge, and also to Dr. Dorothy Needham, whose classic text, Machina Carnis (127), summarizes in a profound and elegant manner a lifetime's study of muscle. For obvious reasons the original writings of the Greek school, and of some of the classical figures subse- quently, are not readily available, and there is the further problem of obtaining translations. In later sections of the chapter the emphasis has been placed on events in the muscles and motoneurons, partly because most is known at this level of the nervous system in relation to exercise, and partly because an analysis of exercise ef- fects in the brain would make the chapter unwieldy. Also, there is the danger that too much attention to the brain would allow the chapter to become not so much a review of exercise as an essay on motor control. At this point it should be added that the re- view is one written by a neurophysiologist and, although muscle metabolism is touched upon, other chapters deal more thoroughly with this topic, as well as with muscle blood flow and oxygen consumption in exercise. On reading the review, some may think that too much attention has been given to the contributions, between the two world wars, of the Oxford, Cambridge and London schools of physiology. It is certainly true that there were excellent centers of exercise physiology elsewhere, no- tably in the Harvard Fatigue Laboratory and in Copenhagen. However, the main re- search directions in these laboratories did not include the excitation and contraction of exercising muscles, and it was in these overlapping fields that the supremacy of the British schools was universally recognized. Even with this caveat, any survey of 39 40 EXERCISE PHYSIOLOGY people and ideas is necessarily selective and so reflects personal choices, although I would like to think that most of my choices would enjoy support. Nevertheless, there are many other muscle physiologists, living and dead, whose work might well have been included in a chapter of this kind and to them I can only apologize. MUSCLES AS THE GENERATORS OF FORCES AND MOVEMENTS Nowadays, it seems that even young children have an awareness that certain soft swellings in their bodies are muscles and that, in some way, the size of a muscle, most commonly the brachial biceps, is an indicator of strength. Yet such knowledge was not always apparent, for Hippocrates (460-380 B.c.) and the Hellenistic school con- sidered that the tendons, rather than the muscle bellies, had the ability to produce movement. Moreover, because of their often similar gross appearances, the tendons were not distinguished from nerves. This confusion is evident in the following de- scription of the functions of the principal tissues, given at the end of the third cen- tury B.C. "The bones give a body support, straightness and form; the nerves [ten- dons] give the power of bending, contraction and extension; the flesh and the skin bind the whole together and confer arrangement on it; the blood-vessels spread throughout the body, supply breath and flux and initiate movement" (85; bracketed interpolation added). For Aristotle (384-322 B.c.) also, the tendons were the body structures responsible for producing movements since, like an automatic puppet, "Animals have parts of a similar kind, their organs, the sinewy tendons to wit and the bones; the bones are like the wooden levers in the automaton, and the iron; the tendons are like the strings, for when these are tightened or released movement be- gins" (3). In contrast to the sinewy tendons, the muscles or "flesh" were thought to convey the sense of touch: "chief of all the primary sensibility is that of touch; and it is the flesh, or analogous substance, which is the organ of this sense" (4). It was Herophilus of the Alexandrian school (early third century B.c.) who ap- pears to have been the first to recognize the involvement of muscles in producing movements and to distinguish between nerves and tendons, as well as between ar- teries and veins. Erasistratus, a contemporary of Herophilus, thought that the muscles contracted because they filled with pneuma, a vital spirit derived from the air. This spirit was conveyed, in an altered form, from the brain to the muscles by means of the hollow tubes of the nerves. It was thought that as the muscles fill with pneuma they "increase in breadth but diminish in length, and for this reason are contracted" (44). It was Galen (129-199 A.D.), however, who took the understanding of muscles and nerves to entirely new heights. As well as carrying out public dissections of human bodies, Galen experimented on live animals, and on African monkeys in par- ticular. He wrote copiously on structure and function, as well as on the practice of medicine; he was also a noted philosopher. From his anatomical descriptions, it is ap- parent that Galen must have been extremely skilled in dissecting human and animal bodies and in recognizing, in great detail, the different parts so revealed. In relation The Neuromuscular System 41 to muscles, he clearly saw that these could only be in one of two states, contracted or relaxed, though he viewed relaxation, incorrectly, as resulting from the pull of an- tagonist muscles: "The natural activity of the muscles consists of contracting and withdrawing upon themselves, and lengthening and relaxation take place when the antagonist muscles pull and draw towards themselves" (52). So formidable was Galen's intellect, and so extensive and complete were his writings, that his views on muscle contraction remained unchallenged until the Renaissance and the appearance of the next major figure, Andreas Vesalius (1514-1564; Fig. 2.1). Born in Brussels, Vesalius dissected animals while still a classics student at the University of Louvain. Given his medical family background and biological curiosity, it was natural that he should become a physician. In pursuit of that goal, he moved to Paris to be taught by Sylvius. Vesalius soon became dissatisfied with the passive recital of Galen's treatises and, in particular, by dissections which were, by his own standards in animals, clum- sily undertaken. Vesalius's evident abilities led the rulers of Venice to appoint him to the chair of surgery and anatomy in the new University of Padua, in which position Fig. 2.1. Andreas Vesalius (1514-1564). A very handsome portrait of the famous anatomist and pioneer physiologist, seen here holding the dissected flexor ten- dons to the fingers. Reproduced from Foster (51), with permission from Cam- bridge University Press. 42 EXERCISE PHYSIOLOGY in it was his responsibility to perform public dissections. This was exactly the op- portunity which Vesalius was seeking. After 5 years of unrelenting endeavor, he pub- lished his observations on the dissected cadavers in a monumental work entitled De Humani Corporis Fabrica (Structure of the Human Body). The extraordinary detail of the anatomical parts was illustrated by many plates and woodcuts, the creation of another Belgian, John Stephen Calcar. When it came to neuromuscular function, Vesalius accepted much of Galen's teaching, as is evident from the following quotation: Muscle therefore, which is the instrument of voluntary movement as the eye is the in- strument of visiorr and the tongue of taste, is composed of the substance of the ligament or tendon divided into a great number of fibres and of flesh containing and embracing those fibres. It also receives branches of arteries, veins and nerves, and by reason of the presence of the nerves is never destitute of animal spirits so long as the animal is sound and well.... I am persuaded that the flesh of muscles, which is different from everything else in the whole body, is the chief agent, by aid of which (the nerves, the messengers of the animal spirits not being wanting) the muscle becomes thicker, shortens and gath- ers itself together, and so draws to itself and moves the part to which it is attached, and by help of which it again relaxes and extends, and so lets go again of the part which it had so drawn. (147) By reviving Galen's practice of animal experimentation, Vesalius pointed a way for others to follow; in this respect, he and Galen can be considered the founders of physiology. By vivisection Vesalius was able to show directly-by the simple expe- dient of ligating a nerve and observing the paralyzing effect on the contractions of the struggling animal-that the muscles really did contract under the influence of nerves. When the ligature was released, the contractions resumed. By dissection he was also able to show that the nerve influence was mediated by the interior of the nerve rather than the surrounding sheath, though he refused to speculate on whether the "animal spirit" was transported in hollow channels within the nerve or through the solid material of the nerves-a dilemma which distantly echoes the present debate over the relative importance of impulse activity and axoplasmic transport in maintaining muscle properties (seep. 76). Vesalius made important observations on other parts of the body and, while he did not challenge some of Galen's doctrines openly, he gave strong hints of his scep- ticism. This tendency to deviate from the orthodox view did not pass unnoticed by those who had read the Fabrica and, perhaps because of envy, there was resentment and political maneuvering against Vesalius. In anger and disappointment Vesalius turned his back on an academic career by accepting a comfortable appointment as court physician to Emperor Charles V. He eventually died during a journey to Venice. A more exact description of muscle contraction did not appear until100 years later, in 1664, with the publication of De Musculis et Glandulis Observationum Specimen by the Dane Nicholas Stensen (1638-1686). Apart from being a skilled anatomist (the discoverer of the parotid duct), Stensen was an early geologist who, in effect, founded crystallography and was perhaps the first to recognize the signifi- cance of fossils as a biological record. In his anatomical studies, Stensen used one of The Neuromuscular System 43 the early microscopes to examine muscle. He described "motor fibers" each of which was composed of the "most minute fibrils" arranged lengthways. From his account it is evident that Stensen's motor fibers were the muscle fascicles and that the most minute fibrils were the muscle fibers themselves. He went on to state that it was only the fleshy parts of the motor fibers which contracted while the tendinous parts, at the ends of the motor fibers, remained unchanged (144). Francis Glisson (1597-1677), while regius professor of physics at Cambridge, was one of those who disputed the notion that the nerves conveyed some spirit to the muscles which inflated them and produced the contractions. He put the matter to experimental test by having a subject insert his arm in a glass tube and making a watertight seal around the shoulder. He was then able to show that, when the sub- ject made a forceful contraction, no water was displaced from the tube into a vertical sidearm-that is, the volume of the contracting muscle remained constant. Impor- tant observations on the nature of muscle contraction were also made by Glisson's contemporary, Giovanni Borelli (1608-1679). An able mathematician, Borelli was born in Naples but subsequently moved to Rome under the patronage of Christina, the former queen of Sweden. A rough man, largely self-taught and possessed of many interests, Borelli investigated by the application of anatomy and physics, the ways in which movements were produced. Through calculations of angles and lever arms, Borelli was able to analyze a variety of movements, including walking and run- ning, in mathematical terms, and much of his work remains valid today. Like Glisson, he refuted the notion of animal spirits inflating the muscle, point- ing out that no air bubbles could ever be seen to come from cut muscles contracting underwater. Instead it was the muscle tissue itself which contained all the material necessary for the contraction. In regard to the action of the nerves in initiating con- tractions, Borelli thought that possibly "some commotion must be communicated along some substance in the nerve, in such a way that a very powerful inflation can be brought about in the twinkling of an eye" (16)-a statement foreshadowing the discovery of the action potential. And so, with the publication of Borelli's treatise on animal motion (De Motu Animalium), the correct relationship of the roles of nerves, muscles, and tendons in exercise was finally established. Insights into the nature of the contractile mecha- nism itself would not be advanced significantly until the development of the sliding filament hypothesis 300 years later (see the section on Machina Carnis: the Flesh Machine). THE MOTOR UNIT That there are probably many more muscle fibers than motor nerve fibers must have been obvious to those at the end of the nineteenth century who were able to cut and stain thin cross sections of nerve. The implication of this discrepancy in numbers was clear to Charles Sherrington (1857-1952; Fig. 2.2), who realized not only that each motor nerve fiber must supply many muscle fibers but also that each nerve fiber and 44 EXERCISE PHYSIOLOGY Fig. 2.2. John (Jack) Eccles (1903-1997) and Sir Charles Sherrington (1857- 1952). This photograph, a well-known one, was taken on one of Eccles' visits to Sher- rington at his Ipswich home after Sherrington's retirement in 1935 and before Ec- cles' return to Australia in 1937. From their postures, it would appear that Sher- rington has just asked Eccles a difficult question. Unfortunately, there are few photographs of Sherrington as a young man but, even in old age, the lively intel- ligence is still very evident in his face. Reproduced from Granit (68) with permis- sion. its colony of muscle fibers could be considered a "motor unit." Having shown that, unlike the situation in crustaceans, mammalian muscles did not have a local in- hibitory nerve supply, Sherrington recognized that, in exercise, all movements and forces must be graded according to the numbers of motor units participating. The numbers of active units would, in turn, depend on the numbers of motoneurons in the spinal cord or brain stem that were discharging under the antagonistic influences of central excitation and inhibition. More than any other concept except for that of the contractile mechanism, the motor unit has dominated muscle physiology and has spawned a set of subsidiary questions each of which has importance in its own right. But, first, how could the sizes and tensions of the motor units be measured in a mammalian muscle? At the time he undertook this problem, in 1928, Sir Charles Scott Sherrington had been Wayneflete professor of physiology at the University of Oxford for 14 The Neuromuscular System 45 years. He was then in his 70th year but as full of energy and enthusiasm as ever, and a magnet for aspiring neurophysiologists from all parts of the world. A shy, unas- suming man, slight in build but still sharp of eye and mind, Sherrington had enjoyed a life of great academic distinction. Knighted, a former president of the Royal Soci- ety, and the recipient of numerous awards and honorary degrees, Sherrington was in the last phase of an extraordinarily successful academic career when he began to ex- plore the motor unit. To count the number of motor nerve fibers, and hence the number of motor units in a muscle, it was first necessary to remove the sensory fibers from the mus- cle nerve. Sherrington did this in the cat by allowing the sensory fibers to degener- ate after expertly dividing the dorsal nerve roots between their ganglia and their junctions with the ventral roots. Despite his age and exalted position, Sherrington did much of the nerve histology himself, taking the specimens home and often work- ing through the night in his bathroom (39). The sections of nerve, stained with osmic acid, were given to a young doctoral student from Melbourne, John (Jack) Eccles (Fig. 2.2). Eccles discovered that there were two populations of motor nerve fibers, large and small, and the total number of fibers was divided into the tensions that had been developed by the whole muscle, so as to derive the mean motor unit force. Eccles and Sherrington thought, however, that the small fibers were those which, during devel- opment, had arrived too late at the muscle to acquire their full complement of mus- cle fibers. It was only shown later, by Lars Leksell in Stockholm, Sweden (106), that the small nerve fibers supplied the muscle spindles. Ironically it was Sherrington himself who had first described the motor innervation of the spindles many years earlier. Nevertheless, through the work of Sherrington and Eccles, the first estimates of motor unit numbers and forces had been made in mammalian muscles and the re- sults, distorted as they were by the inclusion of spindle efferents, showed that there were major differences between muscles. The cat gastrocnemius, for example, had twice as many motor units as the semitendinosus, and the gastrocnemius motor units developed four times the mean force (138). In relation to human muscles, the first quantitative studies were those of Bertram Feinstein and his colleagues in Lund, Sweden, some years later. On the basis of a postmortem study of a patient with poliomyelitis, they assumed that 60% of the large-diameter fibers in a muscle nerve were a-motor. In a variety of limb and cra- nial muscles they found that the mean sizes of the motor units ranged from more than a 1000 muscle fibers in the medial gastrocnemius to 10 or so fibers in the lat- eral rectus of the eye, while the mean numbers of motor units in the same muscles were 579 and 2970, respectively (45). Later still, by recording mean motor-unit po- tential sizes or twitch tensions and comparing these with those of the whole muscle, it became possible to estimate the numbers of human motor units during life (118). FAST AND SLOW MUSCLES AND MOTOR UNITS Throughout history those who ate meat could not have helped noticing that some muscles or parts of muscles were darker than others, but the possible significance of 46 EXERCISE PHYSIOLOGY these variations in color could not be considered until the muscle bellies, and their component fibers, had been identified as the source of contractions (see first section). Perhaps the first serious comment on the color of muscles was that of Nicholas Stensen (1664), who stated: "The middle parts of the motor fibres [muscle fasciculi], wrapped round by the membraneous fibrillae [muscle fibres], constitute together the fleshy part of the muscle, which, soft, broad and thick, differs in colour in different animals, being reddish or pale or even whitish; in the leg of the rabbit you will find some muscles red and others pale" (144). The next step, a major one, came from Louis-Antoine Ranvier (1835-1922), who began his scientific career as an assistant to Claude Bernard before becoming director of the Histology Laboratory and then professor of anatomy at the College de France in Paris. While with Claude Bernard in 1867, Ranvier began to study the fine structure of nerve and muscle and combined his histological investigations with recordings of muscle responses to tetanic stimu- lation. In comparison with pale muscles, Ranvier noted that red muscles contracted slowly, developed fused tetanic contractions at lower rates of stimulation, and were more resistant to fatigue (131). These were remarkable observations for the time and were not equalled until the work of Denny-Brown. Born and educated in New Zealand, Derek Denny-Brown developed a strong in- terest in the nervous system while a medical student, and then as a demonstrator in anatomy (Fig. 2.3, top). To further his knowledge, he joined Sherrington at Oxford in 1925. Although Oxford, because of Sherrington's presence, was the world center for neuroscientific research, the physiology laboratories were primitive: the rooms were cold, damp, and unheated, some lacked electrical outlets, and the doctoral (D.Phil.) students were crowded together. Each laboratory, however, was equipped with a falling plate camera and optical myograph to record the contractile responses. In this system, which required the room to be darkened, the fall of a photographic plate coincided with the release of a stimulus from an induction coil. The ensuing contraction was recorded via the attachment of the muscle tendon to a mirror mounted on a torsion rod, and a narrow beam of light, reflected from the mirror on to the falling plate, registered the tension developed by the muscle (Fig. 2.3, bottom). This sensitive but cumbersome system was not free from error, but in the cat hindlimb Denny-Brown (33) was nevertheless able to confirm Ranvier's observa- tions that, in mammals, the (pale) gastrocnemius muscle had a faster twitch than the (red) soleus; moreover he found that this difference was independent of fiber diam- eter or the amount of fatty granular material in the fibers. The small rectus muscles of the eyeball had twitches which were faster still. Interestingly, the difference be- tween the gastrocnemius and soleus was reversed in the newborn kitten. After Denny-Brown's work showing differences between fast and slow muscles, it would have been logical to study individual motor units and to determine to what extent their properties differed from one another. It is tantalizing to realize that such a study could so readily have been carried out in Oxford, but it was not to be. Instead, the next advance in the fast and slow story was to come from a quite different direction-from the histology laboratory. Wachstein and Meisel may have been the first, in 1955, to show that muscle fibers could be differentiated from each The Neuromuscular System 47 Fig. 2.3. Top: Derek Denny-Brown (1901-1981). Almost to the end of his long ca- reer, he combined clinical neurology with neuropathology and neurophysiology, with conspicuous success in all three. Reproduced from Gilliatt (59), with permis- sion from The Canadian Journal of Neurological Sciences 8:271-273, 1981. Bottom: The first satisfactory photographs to be published of isometric muscle twitches, recorded with the Oxford optical myograph and falling-plate camera. The re- sponses, from a cat medial gastrocnemius muscle, are to paired stimuli of increas- ing separation (A-H). The vertical lines, made by a beam of light emerging from a slit in a rotating drum, represent 20 ms intervals. Reproduced from Cooper and Eccles (30), with permission from The Physiological Society. other by the avidity with which they stained for certain enzyme reaction products (152). After a stain for succinic dehydrogenase, by Wachstein and Meisel, came one for phosphorylase, by Victor Dubowitz and Everson Pearse (36), and then the most useful stain of all, that for myosin ATPase by King Engel (43). All of these and oth- ers, including one for muscle glycogen, revealed the presence of differently staining fibers distributed in an apparently random manner across the muscle belly, giving a 48 EXERCISE PHYSIOLOGY checkerboard appearance. With the myosin ATPase reaction, it was possible to show that there were two main types of fiber (I and II) and that the type II fibers could be subdivided by varying the pH of the incubating medium. Thus was born the science of muscle histochemistry. Rather later came the key observation-that all the mus- cle fibers belonging to a motor unit were of the same histochemical type (40). It was not until 1965 that the physiology of fast and slow muscle moved for- ward, and this was through the work of Elwood Henneman (1915-1996; Fig. 2.4, left). At the time Henneman was a professor of physiology at Harvard. He had orig- inally intended to make a career as a neurosurgeon; in preparation he had studied neurophysiology, and it was probably through David Lloyd's influence at the Rock- efeller Institute that he had become interested in spinal reflexes. However, unlike the earlier situation in Oxford, muscle seems not to have had any priority at Harvard, and Henneman himself, like Sherrington a quiet and unassuming person, appears as a rather solitary figure. Soon after his appointment to Harvard, Henneman became interested in single motor units from the point of view of their recruitment and, in 1957, had his first publication on this subject (see next section). His technique was to Fig. 2.4. Left: Elwood Henneman (1915-1996), who was the first to record the twitch and tetanic responses of single mammalian motor units, and to propose that the order of motor unit recruitment is governed by the sizes of the parent mo- toneurons. Courtesy of Dr. Abby Henneman. Right: Robert Burke (1934-) whose own motor unit studies were influenced by Henneman's work, and who gave the first comprehensive description of the different types of mammalian motor unit. Courtesy of Dr. Burke. The Neuromuscular System 49 dissect out sectioned ventral root filaments and to record the centrifugal impulses of two or more fibers during reflex discharges. Later, Henneman took the dissections further by splitting the filaments distal to the section until only a single axon re- mained on the wire electrode. The isolation of a single fiber was proven by the pres- ence of an ali-or-nothing antidromic response when the motor nerve was stimulated in the periphery. The single motor axon could then be stimulated and the contractile responses of the corresponding motor unit could be recorded from the muscle. In two papers, published back to back in 1965, Henneman, with two younger colleagues, determined the axon conduction velocities, twitch contraction times, and the twitch and tetanic tensions of single motor units in the cat soleus and gastrocne- mius muscles (119, 157). In the red soleus muscle, which they found to be largely composed of one histochemical fiber type, the contraction times tended to be long, and most of the tetanic tension was generated at relatively low stimulus frequencies. Even in this uniformly staining muscle, however, the maximum twitch tensions of the units varied more than 10-fold, indicating a similar range in motor unit sizes. In the pale gastrocnemius muscle, the range of tensions was even higher, 100-fold, and there was a greater range of contraction times too, with shorter ones predominating; these units required higher stimulus frequencies to develop maximum force. These exciting results brought in the modern era of motor unit physiology. Very soon afterward Robert Burke* (an asterisk next to a scientist's name indicates that the scientist is included in the group photograph; see Fig. 2.18) (1934-; Fig. 2.4, right), newly appointed to the National Institutes of Health, Bethesda, carried the physiology one step further by stimulating single motoneurons with intracellular microelectrodes rather than by dissecting out single ventral root fibers (22). In this way he was able to look for any association between the properties of the parent mo- toneuron and those of the axons and muscle fibers; further, by the technique of glycogen depletion, he was able to count the numbers of muscle fibers in a selected motor unit and to show that the fibers tended to be widely dispersed across the mus- cle belly. Finally, by combining glycogen depletion with enzyme histochemistry, he could correlate the physiological and histochemical features of the individual motor units. Burke and his colleagues showed that in cat hindlimb muscles there were three main types of motor units-a large, readily fatiguable unit with a fast twitch; a rather smaller unit, which was more fatigue resistant and had a fast twitch; and a considerably smaller unit, which was difficult to fatigue and had a slow twitch. These three types they termed FF (fast contracting, fast fatigue), FR (fast contracting, fa- tigue resistant), and S (slow contracting), respectively (23). Similar conclusions were reached by V.R. (Reggie) Edgerton* and James Peter, at UCLA (129), on the basis of detailed biochemical studies on rabbit and guinea pig muscles, and by Douglas Stu- art and Roger Enoka in Tucson, who, like Burke, employed single motor-unit tech- niques in the cat. In Edgerton and Peter's classification, the three types of units were termed fast twitch-glycolytic, fast twitch-oxidative glycolytic and slow twitch- oxidative, respectively. In human muscles single motor-unit twitches were first recorded by the technique of threshold motor nerve stimulation and were shown to have a threefold range in contraction times (from 35 to 105 ms; ref. 140; Fig. 2.5, top).  _I\_ "c: "0 0 E 0 ~ "E.::> z" Co ntractio n l ime (ms) (A) (B) 100 z 0 ~ o0o c: ·~c: 0 0 - ~ -\7Qo 0.c 0 ~ 10 ~ 0 0 0 0 I 0.2 20 T hreshold force (N) 50 The Neuromuscular System 51 These results were confirmed and extended by intramuscular nerve fiber stimulation in the gastrocnemius by John Stephens and his colleagues in London; as did Burke in the cat, they found evidence of major differences in motor unit type (56). In more recent studies the main addition to our knowledge of motor units was the recognition that, while certain properties were indeed linked together, as Henne- man and Burke had shown, there was actually a continuum of properties among the motor units. Further, the linkage was not inviolable for, at least in some small mus- cles of the human hand and foot, the relationship between motor unit size and con- traction time was absent. This discrepancy introduces a further word of caution. Ex- ercise physiologists talk of 11 fast" and 11 slow" muscles, by which they mean fast contracting and slow contracting. Ideally, these terms should refer to the rate of myosin cross-bridge cycling and hence to the myosin-ATPase reaction. At a cellular level, the speed of contraction is best measured by observing the rate of sarcomere shortening in an isotonic contraction. However, muscle twitch recordings in the whole animal are invariably isometric and the durations are determined not so much by the rate of the myosin ATPase reaction as by the amount of time that calcium ions are available to the cross bridges before being taken back into the sarcoplasmic retic- ulum. The correlations found between contraction time and other motor unit vari- ables actually depend on the differences in· calcium pumping that exist between the fiber types. THE SIZE PRINCIPLE Having demonstrated the marked variation in motor unit properties, even within the same muscle, it was perhaps inevitable that Henneman should have investigated the Fig. 2.5. Top: First records of single human motor unit twitches, obtained from the extensor hallucis brevis muscle using threshold stimulation, transient arterial occlusion, and response averaging. A "fast" unit twitch is shown at the left and a "slow" unit twitch below and to the right. The histogram gives the twitch con- traction times of 122 units in 31 individuals. Adapted from Sica and McComas (140), Journal of Neurology, Neurosurgery and Psychiatry (1970) 34:113-120 with per- mission from the BMJ Publishing Group. Bottom: Demonstration of the size prin- ciple in the human first dorsal interosseous muscle of the hand. The sizes of the motor units, as reflected in their twitch tensions, are given on the vertical axis, while the horizontal axis depicts the muscle force at which each unit was re- cruited. Filled circles are measurements from single experiments while the open circles are measurements in other experiments with the same subject. The twitch tension of a unit was derived by using the motor unit potential, recorded with an intramuscular needle electrode, to trigger an averager receiving the force output from the whole muscle. With repetition, the force contribution of the triggering motor unit becomes increasingly evident. Reproduced (lower) from Milner- Brown et al. (125), with permission from the authors and from The Physiological Society. 52 EXERCISE PHYSIOLOGY possible ways that these differences might be reflected in usage. He had, in fact, con- sidered this problem earlier by recording the discharges of single ventral root fibers in cats during flexion reflexes (75). In that study he had found that the axons of the motoneurons with the lowest thresholds generated the smallest action potentials. With some support from earlier studies by Eccles and Sherrington, Henneman ar- gued that the most readily activated axons belonged to the smallest motoneurons and that these innervated the smallest colonies of muscle fibers. In passing it is in- structive to ask why there should be any correlation between the respective sizes of the motor axons and motor units. The answer is that every motor axon necessarily contains enough microtubules and neurofilaments not only for its own upkeep, via axoplasmic transport, but also for the upkeep of the muscle fibers which it inner- vates. Hence a large motor unit, with many axonal branches, requires a large num- ber of microtubules and neurofilaments, and a thick axon to house them. But back to Henneman.A problem in his 1957 study was that the motoneuron discharges in the flexor withdrawal reflex were not sustained. There was a burst of impulse activity, lasting perhaps a second, and then silence, so the amplitudes of the nerve fiber action potentials could only be compared over a short period of time. In the later study, with George Somjen and David Carpenter (76), the stretch reflex was employed instead, and in this reflex most motoneurons maintained their discharge for as long as the muscle remained stretched. Using silver wires to record from thin ventral root filaments, the investigators found that motor axons with small action potentials discharged before those with larger potentials, as the reflex became stronger. On the basis of earlier biophysical observations of Herbert Gasser and morphological studies of Ramon y Cajal (26), they argued that motoneuron recruit- ment was organized according to the sizes of the cell bodies-that is, the larger the cell body, the higher the threshold; this was the size principle. It has recently been suggested that priority for formulating the size principle should be given to Denny- Brown on the basis of observations made with Joe Pennybacker during the recruit- ment of human motor units in voluntary contractions (34). In a study of sponta- neous muscle activity in denervated muscle, they wrote: a particular voluntary movement appears to begin always with discharge of the same motor unit. More intense contraction is secured by the addition of more and more units added in a particular sequence. This 'recruitment' of motor units into willed contraction is identical to that occurring in certain reflexes. The early motor units in normal graded voluntary contraction are always in our experience small ones. The larger and more powerful units, each controlling many more muscle fibres, enter contraction late. Interestingly, Denny-Brown and Pennybacker cited Sherrington in relation to the mo- tor unit recruitment that occurs during reflexes and, indeed, Sherrington, in his im- portant 1929 paper summarizing the work of the Oxford school, made it clear that the motoneurons participating in a reflex had differing but reproducible thresholds (138). We know now that Denny-Brown, as in so much, was correct; larger motor units are called into action as the voluntary effort increases. A beautiful illustration of this was to come from the much later work of Richard Stein* and colleagues in The Neuromuscular System 53 Edmonton, Canada (125). In the first dorsal interosseous muscle of the human hand, they used spike-triggered averaging to record "twitch" tensions of single motor units and found that, over a 1000-fold range of tensions, there was a linear relation- ship between motor unit size and threshold (Fig. 2.5, bottom). However, even though Denny-Brown was right, his conclusion rested on flimsy evidence because of the na- ture of his recording electrode. A coaxial needle electrode such as he used can only record from a very few of the muscle fibers in a motor unit at any given location in the muscle belly. Hence, there is a large chance factor in the determination of the am- plitude of the motor unit potential, depending on whether the inner wire of the elec- trode picks up from one, or more than one, fiber-in any case, probably no more than 1% of the full population. But at the time of Denny-Brown and Pennybacker's paper, the correct motor unit architecture was not known and, indeed, the first concept, that the fibers of a healthy motor unit existed in large clumps or "subunits," proved to be entirely wrong. Although Henneman has been criticized for not citing Denny- Brown and Pennybacker's paper (148)-the omission has been described as "partic- ularly perplexing given the fact that Denny-Brown and Henneman were contempo- raries at Harvard"-it is doubtful if Henneman knew of Denny-Brown's work or if the two ever came into meaningful contact in Boston. Further, by the time of Hen- neman's initial report, in 1957, Denny-Brown was more interested in the rostral end of the central nervous system and, as a practising neurologist, in the solution of clin- ical problems. Returning to Henneman's 1965 paper (76), the authors stated that, "we may conclude that there is a general rule or principle applying specifically to motoneu- rones and perhaps to all neurones, according to which the size of a cell determines its threshold." As to the reason for this, Henneman drew on observations on the ampli- tudes of miniature end-plate potentials by Bernard Katz and Stephen Thesleff (100) and concluded that "an equal degree of presynaptic activation may generate a larger synaptic potential in small cells due simply to their dimensions and geometry." Once it had been formulated, and so elegantly, the "size principle" was eagerly adopted by muscle physiologists. It did, after all, make good functional sense, in that weak con- tractions would be smoother and better controlled if they were generated by a num- ber of small units rather than one large unit. However, there was one important dis- senting voice-not concerning the experimental observations underlying the size principle, but the principle itself. The voice belonged to Robert Burke, the investiga- tor who, following Henneman, had carried the correlative analysis of motor unit and motoneuron properties to its ultimate level (see previous section). Using intracellu- lar microelectrodes to stimulate cat medial gastrocnemius motoneurons, he was able to measure the input resistances of the motoneurons and to use these as an indica- tion of their relative sizes, the smallest motoneurons having the highest input resis- tances (21). He found that there was only a weak negative correlation between cell size and the amplitude of the excitatory postsynaptic potential following group Ia (muscle spindle annulospiral afferent) stimulation. Further, he argued, if the densi- ties of the excitatory synaptic contacts on small and large motoneurons were simi- lar, then the synaptic current densities flowing through the membranes of the 54 EXERCISE PHYSIOLOGY respective cell bodies (and axon hillocks) would also be similar despite the disparity in cell sizes. A better explanation for the size phenomenon, he suggested, was that the densities of the excitatory synaptic contacts were higher in the smaller mo- toneurons. Interestingly, Henneman had made the same suggestion, and discarded it, as one of four possible reasons for the size principle: "small cells may receive rela- tively more synaptic input from stretch receptors and cells connected with them." There were a number of observations which supported Burke's proposal, begin- ning with Sherrington's statement that "the very muscles that to the observer are most obviously under excitation by the tonic system are those must obviously in- hibited by the phasic reflex system. (139). While Sherrington's findings showed that the type and/or densities of certain synaptic connections differed among motoneu- ron pools, it could be argued that they did not necessarily indicate differences in synaptic input between motoneurons innervating the same muscle. Evidence of this kind was to come later from experiments both on animals and on human subjects. In one human study, John Stephens in London showed, with colleagues, that the order of motor unit recruitment during voluntary contractions of the first dorsal in- terosseous muscle of the hand could be disrupted if weak electrical stimuli were ap- plied continuously to the adjacent index finger (145). Clearly, then the (polysynap- tic) synaptic connections from the cutaneous afferent fibers were stronger in the motoneurons with higher thresholds for voluntary activation. In a second, and per- haps more dramatic, example, Lennart Grimby* and his coworkers in Stockholm identified some motor units in the small extensor digitorum brevis muscle on the dorsum of the foot that would only discharge during rapid voluntary contractions or sudden corrective movements and might then do so without activation of the units with the lowest thresholds for slower voluntary contractions (71). At a whole- muscle level this last finding was very much in keeping with the old idea of Sher- rington's that different muscles participated preferentially in tonic and phasic con- tractions (see above). In summary, then, a size phenomenon was shown to prevail for slow voluntary contractions and certain reflex responses and, from the above considerations, it would have seemed reasonable to accept that its basis was not the size of the mo- toneuron but rather the type and density of its synaptic connections. This was not the end of the story, however. In 1980 Daniel Kernell and Bert Zwaagstra, in Amsterdam, published the results of a study in which the input resistances of mo- toneurons were compared with the postmortem dimensions of the same neurons after each cell had been marked with procion dye or horseradish peroxide (101). Ker- nell and Zwaagstra, quite unexpectedly, found that the smaller motoneurons had higher specific membrane resistances than the larger cells. The specific membrane re- sistance, it should be explained, is the electrical resistance per unit area of the mem- brane. As Kernell and Zwaagstra pointed out, this meant that, even if all the mo- toneurons in a pool were activated by the same densities of excitatory synapses, the smaller motoneurons, because of their higher specific membrane resistances, would automatically develop larger depolarizations and become more easily recruited than the larger motoneurons. The Neuromuscular System 55 Thus, for a reason which he had not foreseen, Henneman was right after all. Henneman, who died in 1996, was also responsible (with Lome Mendell) for pio- neering spike-triggered averaging, a technique which has had many applications in- cluding, appropriately, the demonstration of the size principle in human interosseous muscles by Stein (see above). THE RECRUITMENT OF DIFFERENT TYPES OF MOTOR UNITS In the preceding section evidence was given that, in slowly developing contractions, and in certain reflexes, the smallest motor units were the first to be called upon. Since studies in animals, also reviewed earlier, had demonstrated correlations be- tween motor unit size and histochemical type, could differential involvement of muscle fiber types be demonstrated in relatively "natural" activities? In the bush- baby, Gala go senegalensis, Edgerton and his colleagues at UCLA had approached this problem by causing the animals to either run on a treadmill or to jump. Histochem- ical analysis of the hindlimb muscles revealed that whereas steady exercise of mod- erate intensity (running) made greatest use of slow twitch-oxidative (type I) fibers, maximal intermittent activity (jumping) depended heavily on fast twitch-glycolytic (type liB) fibers (58). In human subjects, similarly intended experiments could only be performed by analyzing specimens of muscle obtained before, during, and after exercise. The tech- nological advance which made this feasible was the introduction of the muscle biopsy needle by the Swede Jonas Bergstrom in 1962 (9). Perhaps reintroduction would be a better term, for a biopsy needle had been designed and used 100 years earlier by the remarkable French neurologist Guillaume-Benjamin-Amand Duchenne (1806-1875). Duchenne, yet another solitary figure and one regarded as eccentric by his Parisian contemporaries, not only described the type of muscular dystrophy which today bears his name but also was a pioneer electromyographer, a talented medical artist, an inventor of ingenious orthotic devices, and possibly the world's first clinical photographer. His biopsy needle, or "harpoon," consisted of a hollow needle with a side window set back from the tip. When the needle was inserted into the muscle, any fibers bulging through the side window were cut off and retained by a sharp blade pushed sharply down the inside of the needle (37). Bergstrom's design was very similar and his development of needle muscle biopsy was an enormous ad- vance, not only in the field of exercise physiology but also in the diagnosis and as- sessment of neuromuscular disorders. It was fitting that the first application of the Bergstrom needle should have been in Sweden, for a strong school of exercise physiology was developing in Scandinavia, one that could trace its roots back to the work of August Krogh (1874-1949) in Copenhagen. Krogh, a Nobel Laureate and inventor, was a man of wide scientific in- terests but was perhaps best known for his work on the regulation of blood flow in capillaries, including those in contracting frog muscle, and for the diffusion of gases between pulmonary blood and alveolar air. He was also the first to recognize the 56 EXERCISE PHYSIOLOGY exponential decline in oxygen consumption after exercise, a phenomenon later termed "oxygen debt" (see Chapter 8). The studies on muscle fiber-type utilization were carried out in Stockholm by Bengt Saltin * and Philip Gollnick,* who stained biopsy sections of muscle fibers for glycogen and for myosin ATPase activity. The depletion of glycogen detected those fibers which had participated in the exercise while the myosin ATPase identified their fiber type. In cycling, Saltin, Gollnick, and colleagues found that the type I (slow twitch) fibers were most heavily used at sub- maximal workloads, whereas type II (fast twitch) fibers also became involved when the contractions were "supramaximal"-that is, beyond the level at which oxygen consumption was maximal (64). In sustained isometric contractions of the same lat- eral vastus thigh muscles, they found that weak contractions also preferentially em- ployed the type I (slow twitch) fibers and that higher forces made greater use of type II (fast twitch) fibers (63). Thus, both in animals and humans, and in both steady and cyclical exercise, the type I (slow twitch) fibers were the first to be used while type II (fast twitch) fibers contributed increasingly as the effort became greater. FIRING RATE OR RECRUITMENT? In Sherrington's work on motor units it was a central theme that contractions, reflex or voluntary, became stronger through the recruitment of additional units. This re- cruitment came about because more spinal motoneurons had been brought to the threshold for discharging. However, it was also clear to Sherrington, as it had been to Ranvier many years earlier (see p. 46) that force also depended on the frequency of muscle activation as induced by electrical stimulation. To what extent, then, might the frequency of activation vary under natural circumstances? The key to the understanding of this problem depended on the ability to record impulses in the motor nerve or muscle fibers. To record from either tissue, and par- ticularly from the nerve fibers, was not easy in Sherrington's time, but the break- through occurred at the University of Cambridge. In the 1920s the Cambridge Phys- iology Department, still recovering from the effects of World War I, included, as one of its members, Edgar Adrian (1889-1977; Fig. 2.6, top). A charming and gifted man, Adrian nevertheless preferred to work alone in the laboratory, where he enjoyed a reputation for speed and for the knack of getting experiments to work, often with the help of plasticine and pieces of Meccano. Adrian's first success in recording nerve im- pulses came by chance when he had a motor nerve over a pair of recording electrodes and noticed some annoyingly persistent electrical"noise"-which he then realized was the discharges of sensory nerve fibers connected to stretch receptors in the de- pendent muscle belly. Having for the first time "seen" the nerve impulse, Adrian showed that the frequency of the impulse discharge was the information code em- ployed by all the different types of sensory receptor he examined. The next step was to look for a similar code in motor nerve fibers, and he achieved this initially by recording, during reflex contractions in the cat, from fine strands of nerve placed on wire electrodes. For voluntary contractions in human subjects a different approach The Neuromuscular System 57 Fig. 2.6. Top: Lord Adrian (1889-1977) in later life. Adrian became, successively, the head of the Cambridge Physiological Laboratory, president of the Royal Soci- ety, and master of Trinity College, Cambridge. He shared the Nobel Prize with Sherrington in 1932. Reproduced from Zotterman (158), Electroencephalography and Clinical Neurophysiology 44:137-139, 1978 with permission from Elsevier Science. Bottom: Possibly the "cleanest" recording of a single human motor unit discharg- ing during a maximal voluntary contraction. This recording was made with a nee- dle inserted into the adductor pollicis muscle and the unit discharging was one of the few to be innervated by the median nerve, the ulnar nerve having been blocked. The coaxial type of needle electrode was originally designed and used by Adrian and Bronk (1) and is widely used in clinical EMG laboratories. Repro- duced from Marsden et al. (112) with permission from the authors and from The Physiological Society. was needed, and Adrian's solution was simple but effective. By inserting an enam- elled copper wire into a fine hypodermic needle, grinding the end obliquely, and using the wire as one electrode and the needle casing as another, the first coaxial nee- dle recording electrode was created (1). In the triceps brachii muscle he and Detlev Bronk found that during slight contractions action currents appeared at frequencies 58 EXERCISE PHYSIOLOGY as low as 5 Hz and then, with greater effort, reached 50 Hz or more. Adrian recog- nized, and confirmed experimentally in the cat by stimulating the motor nerves, that the higher frequencies would be more effective in developing force. He then showed that the range of impulse frequencies encountered among the motor units during reflex and voluntary contractions corresponded to the rising part of the force:frequency curve. Adrian also demonstrated, however, that the recruitment of additional motor units was an important factor in developing more force. After Adrian, there were many studies of motor unit firing rates in human muscles. In attempts to overcome the problem of deciphering the activities of indi- vidual units from the complex multiunit discharge at the onset of a strong contrac- tion, different types of recording electrode were used, including pairs of fine flexible wires and tungsten microelectrodes. Nevertheless, there was still the problem that the electrode tip would be displaced as the muscle fibers in its vicinity shortened. At Cambridge, many years after Adrian's time, David Marsden, John Meadows, and Patrick Merton (112) took advantage of the anomalous innervation of the adductor pollicis muscle in two of their subjects in whom a very few motor units were derived from the median nerve and the remainder from the ulnar nerve. When the ulnar nerve was blocked by local anesthesia, the discharges of the median-innervated units were easily distinguished in recordings with needle electrodes. The maximum firing rates were found to be as high as 100 Hz or more initially and to decline to 20 Hz or so over the next 30 seconds (Fig. 2.6, bottom). An important addendum to the firing rate issue was that the maintained frequency differed between muscles, as Brenda Bigland-Ritchie* was able to show while working in New Haven, Connecticut (8). With the same type of tungsten microelectrode that had been used previously by others for recording from single human peripheral nerve fibers, she found that the maximum maintained rate for the soleus was only about 11 Hz while for the adduc- tor pollicis and the biceps brachii it was around 30Hz. Bigland-Ritchie and her col- leagues pointed out that this difference in rate was appropriate because in humans, as in other mammals, the soleus had a slow (prolonged) twitch and so developed its maximal tension at low excitation frequencies. But which was the more important strategy for increasing force-rate coding or recruitment of motor units? In going from a series of overlapping twitches at 5 Hz to a fully fused tetanic contraction at 50 Hz, the tension generated by a muscle in- creases 10-fold while there is a 100-fold range in the forces generated by single motor units. In theory, then, both mechanisms could be extremely effective. In their study of the human first dorsal interosseous muscle, already referred to (p. 53), Richard Stein and his colleagues in Edmonton used spike-triggered averaging to dis- criminate the contractions of individual motor units (125). Having determined the twitch tensions of a motor unit, they measured the change in firing rate as the mus- cle contraction became stronger. Using a novel mathematical treatment, they were then able to estimate (124) the relative importance of rate coding and recruitment. In the first dorsal interosseous they found that recruitment was the more important mechanism at low forces, with half of the motor units being engaged when only 10% of the maximum force had been developed. At higher forces, rate coding became in- The Neuromuscular System 59 creasingly important. In other muscles; though, the situation appeared rather differ- ent. Carl Kukulka and Peter Clamann (103), in Richmond, Virginia, showed that while no further motor units were recruited in the adductor pollicis at forces greater than 50% of maximal, in the biceps brachii recruitment continued up to 88% of maximal force. From this and other studies it appeared that the small muscles of the hand depended on rate coding for near-maximal forces, while the larger, more prox- imal, muscles of the arm employed recruitment as well. However, the muscle contractions during which such observations were made were necessarily sustained over many seconds. Were there motor units with still higher thresholds that might only be recruited under different circumstances? In the last section it was seen that at least in the extensor digitorum brevis muscle of the foot there were some motor units which only discharged in very rapid voluntary contractions or in sudden corrective movements. Also, there was, as articulated by the Cambridge physiologist, Patrick Merton, "the belief that lunatics, persons suf- fering from tetanus or convulsions or under hypnosis, and those drowning are ex- ceptionally powerful" (122). Working at the National Hospital for Nervous Diseases in London and usually experimenting on himself with one hand left free to operate the equipment, Merton tested this proposition by the method of twitch occlusion. He designed a ball-race device which enabled the contractions of the adductor pollicis to be isolated when the ulnar nerve was stimulated at the elbow. As the effort to adduct the thumb was increased, the twitch evoked by the interpolated stimulus diminished, almost vanishing when the voluntary contraction was maximal (see p. 58). This diminution indicated that nearly all the motor units in the muscle had already been recruited and were firing impulses at optimal frequencies for force production. Sub- sequent studies have demonstrated that voluntary motor unit activation is complete, or almost so, in most muscles studied, the most notable exceptions being, in some subjects, the triceps surae and the diaphragm. MUSCLE WISDOM In the previous section it was seen that during a maximal voluntary contraction the firing rate of a motoneuron was initially high and then declined to a much lower level. Was there any functional advantage in this? One of the first clues came from a study by David Marsden, John Meadows, and Patrick Merton in Cambridge in which their own adductor pollicis muscles were stimulated tetanically through the ulnar nerves (113). They showed that with a stimulus frequency of 60-80 Hz, fatigue occurred sooner than in a maximal voluntary contraction. However, if the stimulus frequency was progressively lowered from 60 Hz to 20 Hz over the first minute, the rate of fatigue became similar to that in the voluntary contraction. The implication was that in the naturally occurring contraction there was also a progressive reduc- tion in the rate at which the muscle fibers were excited by the motoneurons and that this was optimal for delaying fatigue. Rather provocatively, the Merton group termed this lowering of the muscle firing rate "muscle wisdom." In the course of